Sorry for the delay on this one. Life is still so busy. But hey, here we are with a shorter one. And when the next one hits in May, I swear it will drop at the beginning of the month, and we’ll be back on schedule. For now, enjoy this collection of the most interesting news from the past few weeks!

Let’s dive in.

What Happened the Past Few Weeks

• Skild AI announced a few new partnerships, including one with NVIDIA and Foxconn which will have its “robot brain” powering the assembly of NVIDIA’s flagship Blackwell chips. Currently, this task is done by humans at Foxconn, a factory notorious for its lack of care for humans. So sadly, it’s not astounding that folks are trying to remove humans from the equation entirely. Skild now also has its eye (or rather, A-eye?) on replacing humans in warehouses with its acquisition of Zebra’s robotics division. You may know Zebra for the scanners they make, which are used in most Walmarts to determine how many energy drinks teenagers stole.

• Abridge is now moving into the controversial prior auth space with Highmark Health and AHN. For those unfamiliar with what “prior auth” is, it’s short for “prior authorization” for medical insurance coverage and it’s the reason Luigi Mangione became an alleged murderer. Prior auth is widely viewed as problematic and unethical, as it often results in the delay or denial of care. Many companies, now including Abridge, are trying to develop AI to make the process more real-time, but depending on the implementation, it can actually cause more patients to be unjustly denied care (as in Mangione’s case.) One can only hope that Abridge will do right by patients and providers, but I am sus as always.

• Ok, last bummer I swear. The Pittsburgh International Race Complex was sold for $50 million and will likely be the site of—you guessed it—a new data center. Over 400 acres, including the track were purchased by an LLC with headquarters in Dallas, Texas, a well-known tax shelter. The LLC appears to be a subsidiary of a data center company, as real estate documents were signed by the president of Provident Data Centers. The Post-Gazette piece linked is a sad, but great read, highlighting the perspectives of affected community members. At the state level, the PA House of Representatives just passed a bill to improve data center regulation, but Republicans seem to be on track to kill it in the senate before it can become law, given their current rhetoric.

• Alright, with the bummer stories out of the way, let’s get to some fun stuff. Aurora, Duolingo, and Gather AI all made Fast Company’s 2026 Most Innovative Companies list! Congrats to all of you Aurorans, Duos, and Gatherers. Well-deserved (although the folks that bought Duolingo stock at $540/share last year are still in my prayers.)

• Local startups Meter Feeder and Mapless AI are going to make autonomous vehicles (AVs) pay for parking, just like the rest of us! When an AV pulls up to a curb and shifts into Park, it’ll send a request to Meter Feeder’s API to calculate the cost and pay the city. Good to see our new robo-chauffeurs paying their fair share.

• Speaking of AVs, Waymo is soon rolling out pothole detection in Pittsburgh. The company will make this data available to both the city and Waze users, making it easier for the city to find and repair potholes and for drivers to avoid them.

• PGH Lab, “a six-month paid program for local startups to test new technologies and services with the city government,” has announced their latest cohort. The four startups include technologies such as underwater robots, drones, mapping systems, and (of course) AI. Best of luck to all involved!

• Pittsburgh-based Niche.com released their 2026 Best Places to Live, with Atlantic Station, GA coming in at #1. Places to Live is such a beloved Niche feature, and it’s always fun to see how Pittsburgh neighborhood rankings change each year. My neighborhood of Bellevue came in at #3,107 with our lowest grade being in Crime & Safety. If you’re in Bellevue, please stop doing so much crime so we can get to #1. K, thanks.

• Speaking of #1, last but not least, we had the #1 Trending GitHub Developer, Pittsburgh’s very-own Nate Brake on the blog recently. If you didn’t catch his piece, definitely give it a read or watch the video. He implemented a fantastic idea that puts purpose-built AI in the hands of ordinary folks working in the trades.

For the Days Ahead

• Tomorrow, Pittsburgh Robotics Network will be hosting their Robotics & Tech Happy Hour at The Reserve. Tickets are $20 for tech professionals. Sign up here.

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• This Friday and Saturday, I’ll be volunteering at the 2026 NFL Draft! My assigned role as a wayfinder is on the North Shore. If you see me, please come say hi and feel free to ask where the hot dogs or the nearest bathrooms are. Excited to rep our city and see Wiz perform Friday night!

• May 14th, ProductTank Pittsburgh is doing a meetup about vibecoding. Please, product people, filter your prototypes for slop before handing them to engineers and register here.

• And in case you didn’t get enough Product-type action there, Product Camp Pittsburgh 2026 is Saturday, May 16th at La Roche. I have never been, but I’ve heard great things! Sign up here.

• Do YOU have an event you’d like to share on this very newsletter? Please hit me up!

One Final Note

You may have noticed that I’ve enabled optional paid subscriptions on the blog. As you can probably imagine, even with AI assistance, researching and curating the most interesting local tech news for this blog takes time and effort. It takes even more time and effort for me to write about it, since I never use AI assistance for the writing portion to establish trust with you, the reader, in my authentic voice. At the same time, I’d like to do more with this blog, including more interviews, podcasts, and in-person investigations.

Towards that end, my latest posts will remain free, but posts older than 1 year will now be behind a paywall for folks who want to dig into the archives. If you choose to become a paying subscriber, you will gain immediate access to the archives and be supporting future content.

I love this city and want to see its budding tech community recognized for the absolute gem that it is. So if you’re with me, and you want to see this blog/podcast become an even better source for coverage of the local tech scene, please consider a paid subscription. Thank you, as always, for your support.

That’s it for this month. Have a great week!

—Austin

These days, Pittsburgh is also starting to build a reputation as a global leader in AI, both in its capacity to support data centers and in producing some of the brightest minds in the field, thanks to our world-class universities. As the city and country continue to push AI adoption forwards, I think we’re in a perfect position to consider how the tech can help more than just office workers.

I was recently chatting with Jesse Durham of Wayfinder Contracting, and he shared a story that I think is representative of many who started their own trade business. It goes something like this: they work as a framer or a roofer for a company for a while and one day think to themselves: “Why am I working for this big company when I could go out on my own, do great work, and actually benefit more from the hard work I’m putting in?” So they create an LLC, start doing work, and then quickly realize that running a business is a whole lot more than just being really good at your craft.

There’s all of the administration wrapped around it:

• Visiting job sites to give quotes and estimates

• Researching building material costs

• Planning schedules

• Hiring day laborers or subcontractors

• Sending and managing invoices

• Processing payments

• Filing taxes

This is why a lot of small businesses in the trades struggle or fail. The talent is there. The business experience isn’t. Even when the business is a success, running the business often requires evenings behind the computer managing the paperwork after a long day of doing the physical labor.

Clawbolt.ai: AI that runs the back office for you

That’s why I’ve recently started working on something at mozilla.ai specifically aimed at helping these business owners: Clawbolt.

If you’ve used ChatGPT or general AI assistants before, you know the model: ask a question, get an answer. You can tell it what you need and it’ll accomplish it for you. The problem is when you don’t know what to ask or when you should be asking it. It could schedule calendar reminders if you asked it to, but it’s still sitting in a box waiting for you to log in and type something.

This is where the entire AI ecosystem has started to change, thanks to a revolutionary open source project called OpenClaw that launched late last year. Think of it as a framework for building AI that works in the background on your behalf, not one that only waits for your questions. It’ll take initiative, help let you know about things you didn’t know you needed to handle, and manage things on your behalf.

The catch is that getting OpenClaw set up is insanely difficult, and if you mess something up it carries real security risks. This is exactly where Clawbolt comes in. It’s a very narrow, highly integrated application designed specifically to help contractors and people in the trades who are running their own small businesses and lack the administrative experience or time to handle everything that keeps a business running. Every hour spent typing on a computer sending out estimates and invoices is time not spent on the jobs that are actually earning them their living.

Clawbolt is inspired by the idea of an assistant that will adapt over time to be uniquely yours. It’ll be built with hardened, secure integrations into the core applications and functionality that a small business owner needs, especially someone just starting out. The goal is that they’d interact with it only through messaging apps like Telegram, WhatsApp, or iMessage, and be able to accomplish everything from that single interface. No managing browser sessions, no trying to remember which tab you left open. Clawbolt will be intelligent enough to guide and remind the contractor or business owner about what’s waiting for them, and will help them keep track of everything from calendar items to estimates and beyond.

We know self-hosting isn’t for everyone, so we’re also considering a hosted platform option to make onboarding even easier for people who just want to get started without any of the technical setup.

We want Clawbolt to reduce the stress of administrating a small business and give small business owners back their evenings and weekends.

See It in Action

We Want to Hear From You!

This is where I want to reach out and welcome input from the Pittsburgh community in particular. I live here, and I’m personally invested in this community. I’m pretty confident most of us know a contractor or two who’s drowning in paperwork and spending their evenings staring into a computer instead of relaxing with their family.

If you work in the trades or know someone who does and any of this resonates with you, I’d love to hear your thoughts. Reach out directly at nathan@mozilla.ai or drop a comment below. The earlier we hear from you, the more your input can shape where this goes.

If you’re a software developer, head over to our GitHub project and give it try and a star: https://github.com/mozilla-ai/clawbolt

I would love to see Pittsburgh be the place where AI and the trades come together to bring benefit for us all.

Good morning! I am back from a distant land, recovered from jet lag, and ready to write this post now. The new PIT airport is great though, and contrasts strongly with Chicago’s ORD, most of which feels like a crowded alleyway wrapped in plastic.

Anyways, let’s take a look at what’s been going down in the local tech scene.

What Happened the Past Few Weeks

• The CMU Robotics Innovation Center (RIC) was officially opened in Hazelwood Green (pictured above). Several CMU big wigs, leaders from Hazelwood community organizations, and Governor Josh Shapiro were in attendance for the robot-assisted ribbon-cutting ceremony. The space is huge, and is the target for $1.5 million state grant to fund a “Physical AI Accelerator” to host collaboration among local startups and CMU researchers.

• Shortly after the RIC was opened, CMU announced a big event for the space in partnership with the AI Strike Team: they’re bringing in Mark Cuban for a $1,575,000 sports technology pitch competition the day before the NFL Draft. The event is invite-only (I’ve been poking some of my contacts to see if I can get in, we’ll see.) If you’re interested in pitching, though, you can submit an application on this page.

• FieldAI, a company focused on embodied AI in robotics, is the first firm to inhabit the CMU RIC. They were featured heavily in the opening ceremony, and they appear to have the same general goal as Skild: building a universal robot brain that can operate any sort of hardware. They are also hiring like crazy.

• Speaking of hiring, Gather.ai landed $40M in funding to help with their global expansion, platform development, and scaling their team. They did not disclose their new valuation, but their hiring page is showing some new roles for those who may be interested.

• Aurora continued to expand their driverless trucking operations across the Sun Belt. Aurora driverless trucks are now on the roads between Dallas and Houston, Fort Worth and El Paso, El Paso and Phoenix, Fort Worth and Phoenix, and Dallas and Laredo with a perfect safety record thus far. The company has also updated its software to handle more inclement weather, stating that “inclement weather of all types constrained Aurora’s driverless operations in Texas roughly 40% of the time last year.”

• Technical.ly published updates to their AI data center project tracker page, showing some progress in a few cases. Notably, there doesn’t appear to be any publicly announced progress towards Anthropic’s $1 million partnership with CMU, and a project engineer said electrical work may begin this spring on the $15 billion, 700-acre data centers near Carlisle.

• Just across the state line, West Virginia University Medicine is expanding its use of Pittsburgh-based Abridge’s AI medical transcription systems. The software is saving doctors about 11 minutes per day and “staff who use Abridge also report feeling less stressed or burnt out,” according to Dr. David Rich, Chief Medical Information Officer at WVU Medicine.

• Also in health tech, Xlue, a local startup developing cancer-catching AI, had a great profile piece in Technical.ly. The company has been training its models on health records in Taiwan, and the models are now able to predict cancer with 50-70% accuracy. They are hoping to partner with UPMC soon to gain secure access to more health records with which they can continue to improve accuracy while maintaining patient privacy.

• Meanwhile, up in Troy Hill, startup BioInterphase just moved into to the old Troy Hill firehouse after a beautiful renovation. According to the article, though, the firehouse is rumored to be haunted, and “a medium who the team had come through the space recently said the rumors are true.” The building is apparently inhabited by three protective spirits; hopefully the company can appease them.

• Local web designer guru Brad Frost shared a fantastic video essay on this moment in AI (you can also read it as a text post if you prefer that!) Definitely worth a watch/read if you’ve been trying to process the cognitive whiplash of current events.

• Last but not least, Astrobotic has been awarded a contract by Italy’s space agency to develop a lunar wheel. I’d love to see a lunar Ferrari rocking a set of these:

  

For the Days Ahead

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• This Wednesday, Pittsburgh’s local chapter of getWITit is holding an event entitled Women in Technology: Perspectives Across Career Stages at BNY. The event will be a moderated panel discussion featuring women from many different stages in their tech careers. Sign up here.

• This Thursday’s AI@Work meetup is sold out. This is without question the best local AI meetup geared towards folks actually building things with AI. If you’re interested in joining the community and/or attending the next meetup, sign up here.

• Also this week, there’s a high school robotics competition at the convention center downtown. Admission is free! You can view the schedule here or watch the livestream at the links shown here.

• Coming up next week on March 26th, PyData has another event at CoLab18 called An Unexpected Journey: Life Lessons from Conquering Tesla Solar with Python, a talk about “the unexpected lessons you learn when you trust your instincts and refuse to accept ‘it’s working fine’ as an answer.” Sign up here.

• Just before the NFL Draft, Pittsburgh Robotics Network will be hosting their Robotics & Tech Happy Hour on April 21st at The Reserve. Tickets are $20 for tech professionals. Sign up here.

• If you want an easy way to be notified of local tech events more often than monthly, check out this Luma calendar hosted by Charles Mansfield III. It rocks.

• Do YOU have an event you’d like to share on this very newsletter? Please hit me up!

That’s it for this month. Have a great rest of your week!

—Austin

Good morning! The snow piled up everywhere around the city is finally melting. Now that my pipes have unfrozen, I can finally get around to writing this blog. While everyone may have been staying in during this cold weather, Pittsburgh tech companies were on the move.

Let’s get into it.

What Happened the Past Few Weeks

• Tech stocks have been getting cooked by AI fears, with Duolingo being the biggest loser from its 52-week high based on the above data from Deutsche Bank. A little over a year ago, I reported on how executives at Duolingo were saying “AI” a bunch, then selling all of their shares at an historically high price. In that post, I theorized that the selloff was a tacit acknowledgement from leadership that their stock was significantly overvalued from their “AI” posturing. We will likely never know whether the timing of this selloff was coincidental or intentional, but the market seems to have agreed that the stock was, in fact, overvalued. Could it come back? Is there any way to disable Lucy? Would love to hear your thoughts in the comments!

• Wabtec expanded their operations in eastern India with a new service model that leverages existing Indian Railways infrastructure and staff. The company will be providing maintenance, supervision, material and warehouse management, shed control, logistics, and remote diagnostics to Indian Railways operators. Wabtec also landed a massive order from New York’s MTA for new hybrid locomotives loaded with the latest sensors and safety technology.

• Niche.com quietly laid off a few employees, bringing its headcount back under 300 for the first time in several years. Discovered this initially through posts on LinkedIn and confirmed with sources. After its new college cost comparison tool was featured in the New York Times, the company is seeking to streamline its operations to become more profitable.

• Aurora is now integrated with McLeod Software, allowing more carriers to use its autonomous trucks to handle shipments. McLeod makes Transportation Management System (TMS) software that allows users to manage and book trucking capacity. Now, anyone who has a subscription to Aurora Driver can “tender” (today I learned what this means in trucking, and it’s not referring to delicious fried chicken) shipments in their McLeod TMS to be handled by autonomous trucks, removing the need for separate web portals, etc.

• Technical.ly published some pretty great Pittsburgh-based interview pieces this past month, including one of the City of Pittsburgh’s CIO who resigned after 8 years. Favorite quote from this one was, “Pittsburgh is not able to keep up with the salaries for high-tech IT workers… The further behind they get with that, the harder it becomes to fill those positions.” Max Polec, a well-known local startup coach who is now leaving Pittsburgh, also did an interview in which he clapped back at local critics, saying “I started my first business at 15. I’m 30 now. Do the math.” Finally, local CEO Cara Jones, who is using AI to fight human trafficking at Marinus Analytics, did an interview sharing her story. Her advice to founders: “As long as you’re making forward progress, and you’re making an impact, stay the course.”

• Pittsburgh tech companies had a good showing at this year’s Consumer Electronics Show. According to the Post-Gazette, Strip District-based Aurora, Canonsburg-based Ansys/Synopsys, Bloomfield Robotics in Lawrenceville, North Side-based Bucket Robotics, Hazelwood-based Motional, Squirrel Hill-based Hellbender, and Westinghouse in Cranberry were all in attendance. I would also like to be in attendance next year if anyone has an extra ticket?

• Speaking of the Post-Gazette, on a meta note, I had the chance to attend this Pittsburgh Tomorrow event about the future of Pittsburgh journalism with the Post-Gazette announcing its closure. The event featured a panel of local leaders from the journalism community—WESA, The Trib, Public Source, Pittsburgh Courier, and more represented. I’ll have some more thoughts to share on this in an upcoming post, but looks like a billionaire is already swooping in to buy the Post-Gazette.

• Local open source guru Chad Whitacre shared a post and documentary he released last month on how open source software can help fight tyranny. A great and hopeful watch!

• Last but definitely not least, local physician Desmond Daley published a fantastic opinion piece entitled Pittsburgh’s AI skepticism isn’t ignorance. It’s memory. I’m not going to spoil it here, but it’s a must-read.

For the Days Ahead

• If you’ve been messing with AI lately, you’ve probably heard of Claude Skills. Local AI technologist Ian Cook built a hub for Claude Skills made by Pittsburghers: Skills N’At! You can try other people’s Skills that have already been uploaded or even submit your own on the site. Give it a try!

• Got no plans TONIGHT? There’s an open mic night for local tech founders, researchers, and technologists to pitch companies and share ideas at that new coworking space in Hazelwood Green. Sign up here.

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• Monday, February 23rd, Code & Supply is hosting their recurring Build Night, a time to get together with other folks and build stuff with code at Avenu: Meyran. Sign up here.

• Wednesday, February 25th, PyData Pittsburgh’s first event of 2026 will be Characterizing the Transient Universe with the Pitt-Google Alert Broker. Learn about cutting-edge AI applications in astronomy while exploring one of Pittsburgh’s most fascinating scientific landmarks, the Allegheny Observatory (tours will be offered after the talk)! I was at last year’s talk at the observatory, and I cannot recommend this enough. Sign up here.

• Coming up on March 26th, PyData has another event at CoLab18 called An Unexpected Journey: Life Lessons from Conquering Tesla Solar with Python, a talk about “the unexpected lessons you learn when you trust your instincts and refuse to accept ‘it’s working fine’ as an answer.” Sign up here.

• Do YOU have an event you’d like to share on this very newsletter? Please hit me up!

That’s it for this month. Have a great rest of your week!

—Austin

Good morning and welcome back to the Monthly Beat. This month, I’ll be covering the most interesting pieces of news from both November and December 2025, local events you’ll want to check out in this brand New Year, along with some articles that were featured right here on this very blog!

Let’s take a look at the tail end of 2025.

What Happened over the Holidays

• As Allentown loses its T run and Pittsburghers generally continue to lament the lack of public transportation funding, Waymo has announced the expansion of its driverless taxi service to Pittsburgh. Gleefully appropriating the word “yinz” for their ad copy with a “Yinz ready to ride?”, the company says that several members of their team are already Pittsburgh-based and the city “has a special place in autonomous vehicle history.” While they are currently authorized to operate with a driver onboard in Philly, Waymo has yet to secure a compliance license from PennDOT to operate with or without a driver in Pittsburgh.

• 10 years after its merger, Swedish assistive communication tech company Tobii Dynavox opened a new office in Coraopolis. This was a pretty lowkey headline, but I think it deserves a spotlight here because it’s a great Pittsburgh-centric tech story. Dynavox was an assistive device tech company founded and headquartered here in Pittsburgh that had some financial issues while competing with, well, the iPad. Tobii, a Swedish company that was also competing in the assistive technology space, saw a ton of value in Dynavox, and they offered to acquire it about 10 years ago. Fast-forward to now, the combined company employs over 300 people locally, and they’ve chosen to open a new office here in Pittsburgh rather than move elsewhere.

• Astrobotic secured a $17.5M bag to develop reusable rockets. This infusion of funding “will accelerate development of both its Xodiac and Xogdor platforms, with three distinct vehicles in development.” (Is that… a reference to Trogdor the Burninator with an “X” swapped in for the “T”? Because if so, a multi-million dollar project is named after Trogdor and that’s hilarious.)

• Previous shopping mall and beloved Pirates fan parking spot Nova Place is the latest thing to be marketed as a potential spot for an AI data center. Before anyone jumps to conclusions, I’m just going to quote the Post-Gazette article here: “Jeremy Kronman, vice chairman of CBRE’s Pittsburgh office, said listing the complex as a data center site rather than an office complex is a ‘pricing exercise’ meant to assess the property’s value as demand for data centers surges.” So in my opinion, this is at best a bad joke and at worst bubble behavior.

• Speaking of AI, remember that “AI Strike Team”? They’re rebranding as “Strike Team 2.0” after shaking up their roster of executives. After taking credit for bringing Waymo to Pittsburgh (Side note: Show me the receipts, please. Waymo already had employees here for 10 years. While they seem to feel positively about the team, I’m not exactly sure what the team did.), and posting a few photos in suits, they now have many more actual tech people on the new roster. No one has a silly job title like “Senior AI Strategist” or is talking about “Stargate” anymore, so that’s probably a good sign.

• And now for some quick hits. Synopsys is laying off 10% of its employees after acquiring Ansys. The layoffs are set to occur early this year, and it’s unclear how many of them will be of folks in the Pittsburgh area as of yet.

• Gobble gobble! Just after Thanksgiving, Wabtec acquired Frauscher Sensor Technology Group for €675M (roughly $787.73M). The acquired Austrian company, with 700 employees located in 15 countries, makes railway signaling technologies that Wabtec is eager to incorporate with their other train tech.

• Shortly after scoring a $3M contract to make AI drones for the military, AI company VISIMO filed for chapter 11 bankruptcy. Their reason for filing is perhaps a harbinger for the financial fates of other AI companies in 2026: “Continuous increase in debts”.

• ICYMI, Pittsburgh-local software engineer Sean McPherson shared an opinion piece here on the blog, categorizing varying perspectives on AI. I’ve personally found this chart helpful in navigating conversations about AI with folks who think differently than I do about the technology.

• Last but not least, I reported on two really incredible local events from the past two months, Pittsburgh Robotics & AI Discovery Day and CMU ETC’s Fall Fast. Check out those posts below if you’re interested!

For the Days Ahead

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• The best local AI builders group, AI@Work, is having its 1-year anniversary event January 22nd at 5:30pm. I unfortunately will be out of town for an on-site, but I will absolutely be there in spirit. The past year I’ve spent with this group has helped me not only build out a solid working relationship with AI agents, but also make sense of the whirlwind of change we’ve all been experiencing. Sign up here.

• ProductTank Pittsburgh is hosting a workshop on service blueprinting, January 29th, from 6-8pm at the CMU Integrated Innovation Institute. “Susanna Zlotnikov will introduce what service blueprinting is, why it matters, and how teams can use it to tame complexity, align decisions, and spark innovation opportunities.” Sign up here.

• Do YOU have an event you’d like to share on this very newsletter? Please hit me up!

That’s all for now. Have a great start to your New Year!

—Austin

—Austin

It was a crisp December Friday night—only about 7pm, but already dark enough to feel like midnight—when I had the privilege of attending Carnegie Mellon University Entertainment Technology Center (CMU ETC)’s Fall Festival. The ETC building is just off 2nd Ave across the river from South Side. With its modern, corporate exterior, I didn’t recognize it as my destination at first, but through the glass panes on the side, I could see a flurry of activity on its many floors.

Inside, I was welcomed and given a packet of things to start the night. I won’t lie, I got hype pretty quickly, seeing my class and list of objectives.

I’ll be honest, before this exact moment, I had no idea what this event was going to involve. I hadn’t really even heard of CMU ETC until a friendly DM (thanks Hannah!) with an invitation came my way. As an avid gamer and attendee of many a PAX East, I now knew it was about to be lit.

To start the night, I joined a line for something called “The Curse” designed by the Fall 2025 Experience Design class at CMU. A staff member greeted me and explained that it was an escape room-type experience, so I’d be working with whoever else was in line to solve it.

Only one other person was in line with me, so I struck up a conversation to see if they were mentally prepared to deal with my complete incompetence at escape rooms. Luckily, that person ended up being a super chill local game developer named Izzy who teaches at Highline Coding Academy.

After chatting about all the crazy things going on in the tech world for a bit, we were led inside and given a short tutorial on the mechanics of The Curse. The goal? Break the aforementioned Curse. How does one break such a curse? Well, first you have to solve THIS!

Once we had solved the puzzle on the panel (I won’t spoil), we used a digital book on the panel’s screen to guide us through the rest of the puzzle, which was styled after an Egyptian tomb.

After several minutes of tracing symbols, tapping scarabs, asking for hints from the staff, and poking the pharaohs in the eyes to increase the time limit, we finally broke the titular Curse! Despite my poor performance and need for multiple hints, Izzy and I still had a fantastic time. It blew my mind that this was made by a team of students here in Pittsburgh. The high production quality and attention to detail gave it a great feel, and I could see it starring in any number of escape room businesses.

Short on time, I quickly made my way to the top floor to try more demos. After jealously watching a kid walk a nature trail in a meditative VR hiking experience called Ascent, I sat down with a tablet and played a few rounds of Kicks Lab, an educational game for young sneaker heads.

Kicks Lab was a lot of fun and sported a really cool art style! I designed 5 different shoes for in-game customers according to their needs on a touchscreen tablet. The game simulates the real world physical behavior of various shoe making materials to help players learn how their design choices impact things like comfort and durability in shoes.

Next, I made my way down the hall to a darkened room filled with a massive set of curved projector screens where the 270°Fun team were demoing their interactive F1 racing game. I volunteered to play the “Cheerleader” role and was handed something I didn’t expect: a motion sensor taped to a stick with a flag! Then, as I waved the motion sensor in front of the screen, the in-game crowd started to cheer.

For the next 5 minutes, I was frantically running back and forth to make the crowd cheer so that my teammates could mash a button on their phone screens to juice up the F1 cars as they flew by. Here’s a clip of the victory lap.

With my last few minutes, I chatted with the folks from Team TACIT about their haptic glove VR project, watching as someone virtually ripped a heart out from someone with the glove! This experience is the result of the team experimenting with ways to elicit an emotional response from players with the haptic glove. The concept sounds like it would be super gory and dark, but instead it was surprisingly solemn and emotional—focusing more on the spiritual aspect of letting someone pass away to “the other side.” The cartoon-like graphics help it feel a little less dark as well, with the player physically feeling people’s heartbeats through the haptic glove, then deciding whether to grab the heart and end the person’s life.

I really wish I had budgeted enough time to check out all the other cool stuff on display that night. For next year’s event, I’m blocking off the whole night to try absolutely everything.

—Austin

P.S. If you’re interested in making a game yourself, the first session of Izzy’s upcoming game development class is this Sunday in Regent Square. Sign up here!

A Stage 1 reaction looks something like this: “Ha! Look at this dumb mistake the AI made. It thought I was using Tailwind when I’m actually using Style Components. More like Artificial Unintelligence, am I right? Good luck taking my job, bots.”

A Stage 2 reaction looks more like this: “Hmm, I wonder why the AI got confused there. Maybe some of my classes in this component look like Tailwind classes. How can I prevent this from happening in the future? Let me add a Cursor Rule to tell it I’m actually using Styled Components in this project.”

The naming suggestions a progression. Developers start at the defensive/fearful Stage 1 but are able to progress to Stage 2 with a “curiosity mindset”.

I agree with Ferreira about the two reactions. I feel like I’ve heard them both near verbatim while talking about AI with colleagues. And I agree that curiosity and a growth mindset are invaluable in our field and life broadly. But the framing of “Stage 1/Stage 2” and the motivation behind them is uncharitable (at best) and unhelpful (at worst).^[1]

The difference between the two reactions depends not on time and mindset but on an individual’s belief of AI’s transformative potential and positive impact.

The Skeptical see AI’s capabilities and limitations. They question the imminent advent of AGI and whether future frontier models will have a significant impact on the world in the long-term. They don’t see the technological developments of AI as a net positive. They’ll use phrases like “over-hyped” and “flash in the pan”. They are unlikely to experiment with AI tools. When they see the output of any-given model, they tend to have Reaction 1 because criticism illustrates their beliefs.

The Wary also see AI’s capabilities and limitations but believe that developments in the field will result in dramatic changes to our work and lives. They are concerned that the advancement of this technology will come at severe cost to other aspects of our world. They’ll have a “p(doom)” number, raise concerns about safety, and cite environmental or societal detriments of AI. They may have Reaction 1 but are more likely to avoid AI entirely and encourage others to do the same.

The Pragmatic aren’t swayed by arguments for or projections of AGI. They see AI as a tool to help them with tasks today. They’ll experiment with prompts and models but will quickly move on when it isn’t working. They’ll mention “usefulness”, “utility”, and “productivity”. They spend little time thinking about AI beyond a specific problem and solution. They may have Reaction 2, though there will be a reasonable time box for how long they explore.

The Optimistic see the potential in AI’s capabilities and trust that the limitations will eventually be overcome. They believe that future models will radically transform our world for the better citing problems to be solved in medicine and the environment. They’ll encourage others to learn how to use AI so that they “don’t get left behind”. They tend to have Reaction 2 and will push the boundaries of what is capable with the current state of AI.

I’ve grouped these four perspectives, but everything here is a spectrum. Depending on the context or day, you might find yourself at any point on the graph. And I’ve attempted to describe each perspective generously, because I don’t believe that any are inherently good or bad. I find myself switching between perspectives throughout the day as I implement features, use tools, and read articles. A good team is probably made of members from all perspectives.

Which perspective resonates with you today? Do you also find yourself moving around the graph? I would love to hear your thoughts in the comments, as I’m sure mine will continue to develop over time.

1. Despite my disagreement with Ferreira, I still recommend his newsletter. It has helpful insights and interesting links that I’m happy to see in my inbox. ↩︎

Sean McPherson is a software engineer at Khan Academy building AI-powered learning activities for students. For more, checkout seanmcp.com.

It was a brisk Wednesday morning when I woke up, knocked out an hour or so of work for my day job, then hopped in my car with my dad to head down to the David L. Lawrence Convention Center. We parked on Penn and walked over, stopping briefly at La Prima to get cappuccinos to-go. Once inside, we entered the show floor as soon as it opened, excited to see some local robot action!

Near the entrance, Seegrid’s autonomous forklift bots were already moving crates around. What a relaxing watch! The bot’s motion is so fluid, and it gently comes to a stop well before running into the folks manning the booth. The application here is obvious: a human doesn’t have to lift that crate, either manually or with a forklift.

Next, we stopped by the Advanced Construction Robots Booth where they were showing off TyBot, a robot that ties horizontal rebar intersections​ automatically.

Notice no one is present at the controls. It’s identifying intersections that need tied automatically with visual recognition! This thing can be trusted to do 1,200+ ties/hour day or night, rain or shine.

As we passed by the Thoro booth, their autonomous cleaning bot was doing some circuits while staff talked to onlookers.

We didn’t stand there for long, though, as this super cute lil bot at the HEBI booth caught our eyes.

HEBI not only manufactures robots, but also modular robot parts. Each part of this mobile platform bot can be easily attached to other bots for different applications. As we made our way across the show floor, we came across some other modular HEBI parts at work on a very different robot from the CMU Biorobotics Lab.

This snake-shaped robot is able to traverse and clean nasty, muck-filled storm drains much more easily than a human. And because it’s created with modular parts, it can be extended or shortened as needed for the job at hand.

At this point in our journey through the show floor, a thought that had been slowly forming in my head over the course of the morning took full shape: No one actually wants a humanoid robot—people want specialized robots! The primary goal of building a robot to do a task is to have it complete the task at least as well as a human could, if not better. To achieve that goal, the robot does not have to look like a human, and in fact shouldn’t look like a human, as the humanoid shape is terribly inefficient for robots.

This is a fact known for years by manufacturing and logistics companies, and is the exact reason that the convention center was full of specialized, rather than humanoid, bots. Elon Musk and others who are obsessed with putting humanoid robots in the homes of consumers are spurning the results of years of research and innovation in their pursuit of their idea of the future. In defense of this obsession, they argue that the humanoid aesthetic is what people really want, but the data suggest otherwise.

These headline-grabbing futurists would do well to learn from Pittsburgh-based companies and researchers, who are listening and responding to the actual needs and wants of the market. While there may not have been many robots on display that could be used in the average home, those present at the show make workplaces safer for workers across the board. And based on the data, I’d venture to guess that consumers would much rather have a robot dog walk itself than pay for a humanoid robot walk their real dog!

And the folks at the CMU MATT Lab are making robot dogs smarter than ever.

Last but not least, I do want to shout out the only humanoid robot to be found in the convention center, Agility Robotics’ Digit, pictured at the top of this post. It was sadly supine on the ground when we came by, and its caretakers told us it had malfunctioned and needed to recharge. But it does look pretty neat when in action, and not eerily faceless like some other humanoid bots.

There was also a ton of AI stuff there, but I’ll save that for another post. Till next time!

—Austin

Good morning!

Hope you all are having a pleasantly spooky October so far. Two author’s notes before we jump into this month’s issue:

• I’ll have one or two more long-form articles for you all over the next few weeks, but after that I’m taking a break from writing for the holiday season to recharge. I’ll be back and refreshed in the New Year!

• Starting with this month’s issue, I’m omitting “The Big Story” in favor of posting separate, more in-depth pieces outside this monthly summary. I hope to produce some more unique, locally-sourced content that you all will enjoy. Please let me know any thoughts you might have on the format change!

With that said, let’s get into it.

What’s Been Up This Past Month

• How about a big ole dose of metropolitan news to start us off? Pittsburgh International Airport renovations are finally almost done (94% complete, to be exact), and a lot of new tech is now at play. Skild is deploying robot dogs for security patrols, and Carnegie Mellon robots are scrubbing floors with UV rays. But perhaps most exciting are the new scanners they’ll have at security (pictured above), which are on track to go live right around now, drastically reducing wait times!

• The City of Pittsburgh seems to be getting into the business of social media surveillance. The city plans to start monitoring social media and tie individual users to their comments as part of their expansion of a new Salesforce-based 311 system. I tend to agree with Council Member Charland, who said, “I would rather us stick to picking up the trash and figure out how to do that and get that part right before we talk at all about expanding into any of these kinds of troubling capabilities.”

• The Sheetz on Freeport Road along Interstate 76 (Exit 48) now has a particularly significant EV charging station. It’s PA’s 20th EV charging station, putting the state in 1st place for stations built through the National Electric Vehicle Infrastructure (NEVI) funding program. Since the program’s inception, PA’s stations have delivered 30,000+ charging sessions, powered 3.5M+ miles of driving, and reduced CO2 by 1.7M+ pounds. Sheetz also announced a partnership to deploy even more charging stations across all of its locations in an effort to solidify its position as the favorite haunt of EV drivers.

• Speaking of cars and charging things, Honda and Astrobotic have teamed up to develop power systems to last through nights on the Moon. Astrobotic will be integrating Honda’s regenerative fuel cell system, which stores solar power as hydrogen during the lunar day and converts it to electricity during the 2-week lunar night, with Astrobotic’s own LunaGrid power service and Vertical Solar Array Technology. Whether lunar power will be cheaper than Duquesne Light’s AI data center-related price hikes remains to be seen.

• Rockets and EVs aren’t the only things moving around here, with KEF Robotics moving to East Liberty and pricing analytics company Green Cabbage moving to Warrandale. KEF’s new crib has those SCIFs I mentioned a couple of months back, which they hope to leverage for more Department of Defense work. Both companies plan to hire new employees to fill their respective new spaces.

• While KEF hopes to get more federal funding, CMU was forced to lay off 75 of its Software Engineering Institute (SEI) employees due to Trump Administration funding cuts. SEI is a federally-funded research center focused on AI/cybersecurity for the Department of Defense. This latest move, unrelated to the government shutdown, follows earlier layoffs of over a dozen computer science staff in August, which were also due to Trump Admin funding cuts.

• In AI news, conversation turned to AI brainrot at the locally-hosted Eradicate Hate Global Summit. The summit has been held every year since the Tree of Life synagogue shooting, which claimed the lives of 11 people in 2018 in the deadliest antisemitic attack in our nation’s history. Public Source’s piece shares how one particular panel at the summit discussed “how AI is being used to promote extremism and violence — and how it can be used to stop it.”

• Commentary was also buzzing around AI investment in Pittsburgh this past month. Carta shared data showing that 85.5% of software startup capital in Pittsburgh went to AI, which is the highest in the nation. Meanwhile, NEXTPittsburgh published a great piece in the wake of the AI Horizons summit, questioning if there will be any actual benefit to Pittsburgh for becoming an AI hub. Last but not least, at Technical.ly’s RealLIST Connectors event, leaders said Pittsburgh should focus on physical AI (hardware/robots), not just software.

• Ok, last AI thing I swear—if you’re a software engineer, SRE/DevOps, AI builder, or just one of my more technical readers and haven’t seen it yet, go check out my friend and fellow Pittsburgher Robbie McKinstry’s guest post from this past month on agentic site reliability engineering! It’s such a good breakdown of what’s on the market and what people should be aiming for with these tools. Link below.

• Now for some quick hits. Niche published their 2026 Best Schools and Districts in America Rankings. Three Pittsburgh districts cracked the Top 100: North Allegheny (#25), Mt. Lebanon (#68), and Fox Chapel (#69) (nice).

• Technical.ly posted a great analysis piece highlighting the disconnect between job posting numbers and the difficult reality of finding a tech job locally. Hang in there, folks.

• Schell Games launched a new competitive skydiving game called Project Freefall. Players fight for a sole parachute before hitting the ground, and you can have up to 8 players at one time, either on PC or VR.

• Also in games, Mega Cat Studios founder James Deighan shared answers to 10 questions on Pittsburgh Startup News. Had no idea they now have over 800 remote team members! Wild.

• Lastly, Wabtec is sending its trains and train technology to Kazakhstan. The $4.2B order from the country is the largest in company history and aims to strengthen Kazakhstan as a transit hub between Europe and Asia.

For the Days Ahead

• TONIGHT at 6pm! PyData Pittsburgh is hosting their October event, Driving Materials Innovation with Data: The MDS-Rely Center for Industry-Academic Partnerships. Sign up here.

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• The next Code & Supply Build Night is October 27th at 6pm. Build Night is an impromptu co-working space where you can work on a project, pair, learn, and mingle. Sign up here.

• Pittsburgh Tech Week is November 3rd to 8th. The website is still pretty badly designed, but you can view the events for that week and sign up to sponsor or host an event there.

• Probably the most exciting of that week’s events is Pittsburgh Robotics Network’s Robotics & AI Discovery Day 2025 on November 5th at the David L. Lawrence Convention Center. This year’s discovery day will be focused on the integration of AI into robotics systems. Register to attend here.

• Feeling like you want to take a break from things around the holidays, too? Schedule your time away with Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

As mentioned previously, keep an eye out for one or two more posts before my holiday hiatus. Till then, have a great autumn!

—Austin

Agentic Site Reliability Engineering (SRE) is the latest craze proliferating through Silicon Valley's AI technosphere. The Valley's deepest venture capital pockets are ponying up hundreds of millions to chase this new AI agent use case.

Both Traversal AI and Ciroos AI recently emerged from stealth to throw their hat in the Agentic SRE ring, having raised $48M and $21M Series A rounds respectively. They join the ranks of about ten other startups like Antimetal, Parity, and Neubird, as well as tech blue chips like Datadog and Microsoft.

But what is Agentic SRE, and why has it emerged as one of the hottest AI investments in an already bullish market? If you ask these startups, you won't get a straight answer.

What Do SREs Actually Do?

Site Reliability Engineers (SREs) are the engineers responsible for software uptime. The application in question doesn't matter; whether a mobile app, a web backend, or an IoT fleet, it's the SREs who make sure production is humming along.

They "keep the lights on" for the business, so to speak, dealing with production releases, observability, application instability, and incident triage and response. They also build tools to ensure the lights never go off in the first place.

The term originated with Google in the early 2000s but was popularized in 2016 with the release of the Site Reliability Engineering O'Reilly book. The book positions SRE as an evolution of the DevOps role—where DevOps engineers might be relegated to keeping the lights on and nothing more, SREs are empowered to advance the platform, proactively improving application stability and observability. “Proactive” in this sense is a natural follow-on to the broader “shift-left” mentality in the industry, where responsibility for quality and reliability is shifted earlier in the development cycle.

Ultimately, I submit that the SRE role can be broken down into two broad job functions:

• Reactive work: responding to alerts, mitigating incidents, conducting postmortems, and documenting the results in playbooks or runbooks.

• Proactive work: improving observability tooling, orchestrating canary deployments, advancing application stability, and capacity planning.

Where Agentic SREs Fall Short

With this in mind, you might be tempted to think that Agentic SREs (AI agents that are meant to fulfill the role of an SRE) also have a dual mandate: keep the application running and prevent future incidents. At the very least, you'd expect the AI agent to resolve incidents, autonomously scan logs, match them to runbooks, and mitigate an outage before the human operator gets paged.

In reality, most of these tools do little more than write your Elasticsearch queries for you. Traversal, for example, explicitly advertises that its agent only has read-only access to your production metrics and logs. Similarly, Neubird’s flagship product Hawkeye diagnoses outage root causes but stops short of taking action. Bits AI SRE by Datadog and Incident.io's AI SRE do the same.

Curiously, these agentic SREs are exclusively focused on the reactive SRE job function (i.e. incident response). Beyond that, they're nearly all nonoperational in that they don't perform the tasks required to resolve the incident.

An engineer shelling out $415/month for PagerDuty's AI SRE can expect to respond to the same number of pages as before. The only difference: they'll now have a summary of the incident and suggestions for remediation waiting when they log on. That's certainly not nothing, but it's a far cry from replacing the human SRE in the loop.

Three Limitations of Today’s AI SREs

So why is it that the vast majority of AI SREs are neither operational nor proactive? There are three primary reasons.

1. RAG.

The first reason is historical: AI SREs sprang up because of the sudden accessibility of RAG. With RAG, the agent provider can ingest your playbooks and log data, vectorize them, and toss them into a relational database.

When an alarm sounds, an LLM can generate queries to fetch the most relevant vectorized logs and metric data. You can then prompt it to analyze the output of those queries and hypothesize one or more root causes. The RAG index query serves as a high-pass filter, while the LLM serves as the low-pass filter to remove noise from the observability data.

Because these systems depend on that specific technological advancement, they're limited by what RAG needs to function. Without observability data, the AI can't "suppose" where the next incident will come from, so it can't proactively suggest improvements. This is why all existing AI SREs are reactive alone.

2. They're fighting yesterday's war.

When an AI sees an incident for the first time, and it's going in blind, it has little chance of diagnosing the root cause from the infinite number of possibilities.

Seasoned SREs know a large part of the job is keeping the context of the entire weave of microservices—including the history of its development and interaction—in their heads while resolving an incident. The job requires a lot of tribal knowledge, which is why they document that knowledge in playbooks and postmortems.

An agent can ingest human-authored playbooks to equip itself to diagnose known incident types, but it's the unknown and unexpected incidents that require tacit knowledge to resolve. When a code change creates a new type of incident, only the human has the necessary context to filter the signal from the noise.

As a result, AI SREs will only be capable of resolving known and well-understood incident types. They would be ineffective at taking operational actions to resolve new incidents. To avoid false positives, current AI SREs defer to humans in all cases rather than risk overreaching.

3. Operational agents are dangerous.

Finally, a reactive, operational agent is simply a terrible idea.

Picture this: it's 5:00 AM and your pager goes off. There's an outage in production. You roll out of bed and consider putting on a pot of coffee. Looking at your phone, you see the agent has been thrashing against this incident for the better part of an hour.

By the time you log on, the AI has restarted pods, tainted and replaced cluster nodes, rolled SSL certs, and flushed the logs. Any chance you had to determine the root cause has long since passed.

An operational agent with the keys to production that's able to independently resolve issues with 95% accuracy will wreak havoc 5% of the time. And when things go wrong, it will always be catastrophic.

It's like watching a Roomba glide over the chalk outline at a crime scene: you lose all ability to detect what caused the issue, and you'll be trying to fix one problem while the agent is causing three more. The last thing any SRE wants is to have an AI running amok in production.

The Better Path Forward

This is why agentic SREs are missing the mark. They're focusing on the wrong job functions of the SRE.

Instead, these AI agents should be focusing on proactive, operational SRE tasks: monitoring canary deployments and rolling back before an incident is called, improving distributed trace capture with richer metadata, or fine-tuning alert thresholds to reduce false positives and false negatives. After all, an ounce of prevention is worth a pound of cure.

Consider canary deployments as an example of an alternative. If you use an open source model to run your AI agent on GitHub Actions, your agent can monitor your canary deployment for 180x cheaper than having a DevOps engineer on log-watching duty (using Glassdoor average salary data as a point of reference).

Ideally, you want to equip the agent with tasks that are safe to execute independently. Rolling back a stateless deployment is a safe task. If the agent mistakes a good deployment for a bad one and initiates a rollback, you simply return to what was already running in production. No harm done.

To Wrap It Up

All in all, the current cohort of AI SREs have a long way to go before they can scratch even half of the dual mandate of the SRE. This round of products misses the mark, but not all hope is lost. There are plenty of good opportunities for AI SREs to proactively operationalize the more manual labor of the SRE job.

Good morning!

It’s fall y’all, and that means Steelers football is back baby. Let’s hope Aaron Rodgers doesn’t break or tear something, and we can actually make it past the first round of the playoffs this year. Yesterday’s game was pretty disappointing aside from Jaylen Warren jumping over a dude.

The local tech scene, on the other hand, did not disappoint this past month. Due to my posting this about halfway through September, I’m including some early September stories here in addition to August.

Here’s the latest.

What Happened This Past Month

The Big Story

Local Tech Job Market Worsens, Especially for New Entrants

It’s no secret that the job market has worsened considerably at the national level since the last time I focused on jobs with the Monthly Beat.

Contrary to popular belief, there’s no evidence that AI is affecting the job market quite yet. People can’t seem to agree why, and there has already been quite a lot written about this. So to do something more practical than rehash what’s already out there, for this month, if a company I’m reporting on has open roles, I’ve linked them at the end of each bullet.

Best of luck to everyone out there looking for employment!

The Headlines

• As you may have guessed from the photo up top, Wabtec is now officially partnered with the Steelers. Next time you’re at Heinz Field (no, we’re still not calling it Acrisure) you’ll be able to check out interactive displays showing off train tech, as well as a bunch of Wabtec branding throughout the stadium. Wabtec is also sponsoring various community programs through the partnership. It’s unclear from my research if we will ever see the Pittsburgh Steelers concept train engine in real life. We can only hope. Wabtec has open roles here.

• Also in partnership news, Gather AI inked a new partnership with warehouse equipment company Burwell Material Handling to distribute their drone and camera-based warehouse management products at 20 locations across the U.S. This is definitely a huge milestone for the company, especially considering it’s still based entirely in Pittsburgh and hasn’t migrated any of its operations over to Silicon Valley. Gather has a few open roles here.

• Allegheny County published their first official generative AI policy, which applies to all employees of the county. The policy contains directives like “Users must disclose that GenAI was used to generate content.” and “Users are required to confer with the identified Data Steward to ensure they are using the information in the best way.” If you happen to be the Allegheny County Data Steward or know who that is, drop me a line! I’d love to chat about what that role entails. There are a few county tech jobs here, located under the “Computer Services - System Support” and “Information Technology and Security” categories.

• Everyone locally is starting to get sus of AI data centers, and for good reason. Public Source published a great piece highlighting the impending workforce gap in Western Pennsylvania and how new data centers doesn’t necessarily mean new jobs for people in the community. Meanwhile, folks over on the Pittsburgh subreddit called out that increased Duquesne Light energy prices are a direct result of the advent of such data centers, and much discussion ensued. I am myself sus for multiple reasons, but that’s a topic for another post. Northern Data Group has 3 data center jobs open in Pittsburgh.

• Super Powers Mobility, a New Kensington-based startup that is developing a drop-in electric engine replacement for off-road vehicles, had a great profile piece published in Technical.ly this past month. You can preorder the engine kit and a vehicle that already has it installed on their website, if you’re interested in off-road activities.

• Speaking of Technical.ly, I had the opportunity this past week to meet some of their Pittsburgh RealLIST Innovators and Connectors last week, the former of which was published this past month. Both lists contain great summaries of what these folks are up to and links to their work. Some highlights from the event for me were learning more about local fashion tech startup Olive and music tech startup Lead-In Record Co from their respective founders.

• Now for some quick hits. Gecko Robotics CEO Jake Loosararian posted his take on why the vast majority of generative AI pilots are failing to generate ROI. Gecko has several open roles here.

• Astrobotic is going to be launching their rockets over in Norway. First launch campaigns will start in 2026. Open roles here.

• Korion Health redesigned their website and now have a trademarked name for their smart stethoscope: SoundHeart™. Looks nice!

• Niche published their 2026 Best Colleges Rankings. MIT is back on top. 9 open jobs here.

• Last but not least, at the state level, Pennsylvania’s internet connection plan is a joke, providing only expensive satellite connections to rural folks and claiming those connections are equivalent to full fiber connections. I’ve already written to the Pennsylvania Broadband Development Authority about this and plan to publish a separate article. But if you have friends or family in rural areas outside the city, this is definitely work a read, and I’d encourage you to write to the Authority and your elected officials as well. 13 tech jobs with the state open here.

For the Days Ahead

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• getWITit (Women In Tech Pittsburgh) is hosting a breakfast and panel exploring “the realities, challenges, and successes of being a woman in today’s tech world” this Wednesday from 8am-noon. Get tickets here.

• Decoding Spatial Biology with Python: Multi-Modal Insights into Breast Cancer Progression is PyData Pittsburgh’s next meetup, which will be on the evening of Tuesday, September 30th. Sign up here.

• Spaceburgh is coming up on Sunday, October 5th at Riverview Park, featuring local robotics firms and tons of space stuff. Earlier this year, I had the privilege of touring the Allegheny Observatory, and it was incredible. They’ll be doing tours as part of this event, so I highly recommend checking it out. More details here.

• Pittsburgh Tech Week is November 3rd to 8th. The website is still pretty badly designed, but you can view the events for that week and sign up to sponsor or host an event there.

• Probably the most exciting of that week’s events is Pittsburgh Robotics Network’s Robotics & AI Discovery Day 2025 on November 5th at the David L. Lawrence Convention Center. This year’s discovery day will be focused on the integration of AI into robotics systems. Register to attend here.

  Want to cut aside some time to attend these events? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

That’s all for now, but a deep dive local perspective on a particular breed of AI agents will be hitting your inbox sometime soon. Till then, enjoy a seasonal beverage!

—Austin

Good morning!

Summer is flying by. Hope you are staying cool and hydrated. As I write this, I’m sitting in our dining room near the window with the whirring sound of cicadas and locusts drifting in, reminding me of the whirring of the many servers, sensors, and machines powering Pittsburgh tech companies.

Ok, maybe that’s a stretch, but you get the picture. These bugs are so noisy.

Anyways, let’s get into it.

What Happened This Past Month

The Big Story

The Biggest Investment Announced at the Energy and Innovation Summit is Likely Undisclosed

The city’s top headline was tech-related this past month. The Pennsylvania Energy and Innovation Summit was held here in Pittsburgh, and much of the discussion therein centered around the energy infrastructure necessary to power AI data centers. Donald Trump’s visit for the summit sparked a wave of protests around Carnegie Mellon University (CMU), where it was held, while CMU professors voiced concerns about the summit’s focus on fossil fuels. Indeed, silly partisan quotes such as “You need the natural gas or coal infrastructure in order to provide these giant AI data centers the power that they need.” were plentiful despite the fact that new clean energy infrastructure is cheaper than new fossil fuel infrastructure.

When all was said and done, at first glance, it doesn’t really seem like the summit resulted in anything new or substantive for Pittsburgh outside of CMU. The vast majority of investments promised seem to be for projects in other parts of the state, several of which were already in progress before the summit occurred. For instance, the biggest public investment in Pittsburgh that was touted at the summit, $135 million for an Energy Innovation Center Infrastructure Academy, was actually made over a year ago by the Biden Administration. Additionally, in a few cases, the promised spending announced by Senator McCormick’s office doesn’t actually match the pledges made by the companies themselves.

However, one undisclosed investment, which manifests itself as one, tiny dot just north of Allegheny County on Technical.ly’s funding map, could be game-changing.

At the summit, Westinghouse announced “a plan to deliver 10 AP1000® reactors in the US, with construction to begin by 2030” that will “drive $75 billion of economic value across the United States with $6 billion of value here in Pennsylvania.” Neither of those numbers are the actual amount being invested by the U.S. government, though.

So how much money is this, really? Institute For Progress (IFP), a non-partisan think tank focused on innovation policy, offers a detailed breakdown of the issues contributing to the costs of building reactors. This was a pretty interesting read, but here’s one relevant anecdote they shared at the top:

Two reactors being built in Georgia (the only current nuclear reactors under construction in the U.S.) are projected to cost twice their initial estimates, and two South Carolina reactors were canceled after costs rose from $9.8 billion to $25 billion.

IFP published this article in early May 2023. The final cost of those two reactors in Georgia at the end of May was $31 billion, $3 billion higher than IFP reported at the beginning of the month! So if Westinghouse wants to build 10 new reactors—in the absence of unprecedented zoning reform and other efforts to streamline the process—we’re talking, with some napkin math, at least $155 billion in investment from the U.S. government in a Pittsburgh-based tech company. Obviously, Westinghouse isn’t as trendy a tech company as a new school AI firm, but that’s more funding than all the other announced “investments” combined!

But then again, of those 10 reactors, those built in PA could cost less thanks to, well, unprecedented zoning reform. Governor Josh Shapiro, who also attended the summit, has a track record of cutting red tape to unblock big projects and reduce their costs.

The Headlines

• Skild AI (which makes the robot featured at the top of this post) released a launch video for its “robot brain,” which is able to power several different types of robots. Personally, I wasn’t super impressed by it, as several of the clips in the video appear to be sped up dramatically, and the more difficult tasks that are shown appear to be in highly staged environments. An article from Technical.ly also pointed out that the company’s latest SEC filing doesn’t match up with the funding it purportedly has secured. Would love to hear your thoughts in the comments, though, especially if you have a background in robotics!

• X-Hab has successfully used its 3D concrete printing (3DCP) system to build a structure in subarctic Alaska. This was merely a test structure, but is still the northernmost 3DCP structure on the continent and the first step towards building affordable housing in Alaska with 3DCP (not to mention it’s freaking cool). The company says, “Next spring, the team plans to print a 1,200-square-foot house in the City of Nome based on lessons learned from the test structure.”

• Aurora is now trucking autonomously all through the night. The press release calls out that “37% of fatal crashes involving large trucks occur at night” and that “the Aurora Driver can detect objects in the dark more than 450 meters away, identifying pedestrians, vehicles, and debris up to 11-seconds sooner than a traditional driver.”

• Synopsys finally finished acquiring Ansys. So… should I still report on Ansys? I mean, it’s still here in Pittsburgh, but is it truly Pittsburgh-based now? I don’t know, what do yinz think?

• Folks on the Pittsburgh Subreddit continued to lament the difficult tech job market. Some users debated the root cause, while one user declared, “Part of the problem is no matter how much Pittsburgh wants to pretend it’s some big tech center, it’s just not.” Okay. Obviously, we still have some work to do, as called out by the CEO of Gecko. But I’m pretty sure that’s not even part of the jobs problem, and the problem really is our economy as a whole is slowing down. And if this blog isn’t enough to convince you that Pittsburgh is a “tech center”, I don’t know what is. I won’t say that it’s “big” though—that’s fair.

• For those of you who do mobile development, Carter Levin from Duolingo published an interesting technical article on how they use Swift Macros for iOS development. “Swift Macros is new and not widely adopted yet. But we love trying emerging tools when they make our code cleaner and our team faster,” he says. Duolingo was also one of two Pittsburgh companies that made it onto Time Magazine’s 100 Most Influential Companies list. The other was Dick’s Sporting Goods.

• ICYMI, I posted an opinion piece in reaction to the GPT-5 announcement. This one really seemed to resonate with a lot of folks in multiple ways, and I’ve had so many great conversations as a result.

• Last but not least, for a nice watch/listen as you start your day, InnovatePGH published an interview with local startup founder Miles Mufaka Martin. His startup, Relai, creates and maintains “exchange zones” designed to enable the secure, asynchronous exchange of goods between people.

For the Days Ahead

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• This month’s AI@Work event on the 19th is sold out, but if you’re an AI Builder and interested in joining, you can sign up here. I learn so much at every single one of these. Cannot recommend enough.

• Code & Supply Build Night is Monday, August 25th. Build Night is an impromptu co-working space where you can work on a project, pair, learn, and mingle. Doors open at 6pm. Sign up here.

• PyData Pittsburgh is hosting their next workshop on Thursday, August 28th, entitled “Python for All: Democratizing Coding Mastery with AI Chatbot Support”. Read all about it and sign up here.

• Are you having trouble making time to visit subarctic Alaska to see X-Hab’s 3DCP structures, or even just to get a haircut? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

Next month’s post should be closer to the beginning of the month now that summer’s winding down and I have more time to write. And I should have another bonus piece for you next month as well. Till then, all the best!

—Austin

It followed the usual product launch formula:

• A new version of ChatGPT (GPT-5) was formally announced that will deprecate all previous models.

• Several benchmarks were shared showing that the model scored better than previous OpenAI models.

• Several demos were presented of the model responding to various prompts, mostly focused around front end/web programming.

There was one moment that stood out from the rest of the presentation, though. Sam Altman brought a patient on stage to talk about their journey with cancer and how they used ChatGPT to decide what to do for their medical treatment. The patient shared that their condition was complicated, and that because there wasn’t one specific way to treat it, doctors left the decision to the patient. So, the patient dumped all of their medical forms into ChatGPT and had it make the decision for them.

Did the treatment work, and is the patient getting better? This information was not shared. Did the doctors agree with the decision? This, too, was not shared. Instead, the patient just shared at length how using ChatGPT made the decision feel better.

This is, to say the least, incredibly problematic. Talking to an LLM instead of a doctor to make a medical decision assumes that you have the same level of medical understanding as a trained medical professional. Your doctor has years of medical training and experience, and yes, they could be using the same LLM model you may use to help analyze your medical data. But they are also extensively trained to interpret medical data, whereas you are not. They will know how to prompt an LLM with your medical information better than you. They will know how to use its output better than you. Therefore, if your doctor isn’t helpful in making a medical decision, you shouldn’t be going to ChatGPT for an answer—you should be getting a new doctor!

Beyond this obvious fallacy, it’s hard not to see this part of the presentation as blatant emotional manipulation on the part of OpenAI. This emotional manipulation is not a one-off occurrence—it is at the core of their business and product. Earlier in the stream, the presenters say that the model is OpenAI’s most "emotionally resonant" and "genuine" model yet. They also briefly show a slide that admits the model is deceptive, but the slide is itself deceptive!

If you didn’t catch it, the leftmost bar shows that GPT-5 (with thinking) is 2.6% more deceptive than the earlier model, o3, but the size of the bars is completely disproportionate! Deception and emotional resonance are not qualities you want when seeking medical advice, writing code, or doing anything other than basic conversation and writing generation with an LLM (and even then, you still don’t want deception). Yet these inherent qualities are the default setting in GPT-5 for every user, and can appear in its generated output even after extensive context engineering by a power user. But OpenAI continues to insist that emotional resonance is something we need and that being deceived 50% of the time isn’t that bad, even in dire health situations where an honest, expert opinion is far more important than your feelings.

Don’t be deceived. Emotional manipulation is a feature, not a bug, of OpenAI.

What thoughts do you have about GPT-5? Would love to hear them in the comments. The Monthly Beat will be coming your way next week. Till then, have a great weekend!

—Austin

Good morning!

Hope you all had a great holiday weekend and ate many barbecued foods. I want to say a huge thank you to the 100+ of you who just subscribed in the past month! It was absolutely incredible to be featured in both NEXTPittsburgh and Technical.ly’s newsletter over the past couple of weeks, and I am beyond thankful for the support.

Finally, for those of you who have been here for a while—thank you for sticking around and sharing the blog with friends and family! I really never could have imagined this thing would grow so much, and it’s all because of readers like you spreading the word. For my part, I’ll continue to do my best to bring you 1–2 posts per month of the best Pittsburgh tech content I can muster.

Now then, let’s get into it.

What Happened This Past Month

The Big Story

First Trucks, Now Helicopters—Pittsburgh Tech Companies Push Forward with Autonomous Vehicles

Earlier this year, I reported on Aurora’s historic launch of driverless semi-trucks in Texas. If you’ve been following along, shortly afterward, Aurora had to put drivers back in the trucks because “of certain prototype parts in their base vehicle platform.” This is both exceptionally vague and thoroughly dissatisfying, but despite this setback, the company still made a huge splash with this launch, drawing the attention of the New York Times and other national outlets.

However, it doesn’t seem like national news has yet caught wind of what is an equally impressive achievement by local autonomous flight company Near Earth Autonomy. This past month, the company announced that it had successfully flown a standard Leonardo AW139 helicopter, which would normally require a crew, completely autonomously. While the company still needs to add some more features before this can be used in the field, this is certainly a huge milestone, demonstrating that Near Earth’s autonomous flight systems can not only work in new drones and such that are designed for autonomy, but also in legacy aircraft. The test was performed as part of the U.S. Marine Corps’ Aerial Logistics Connector (ALC) program, and it’s likely the military will be using these systems long before they’re available for commercial helicopters.

The Headlines

• Astrobotic’s new rover (pictured at the top of the post) passed testing and is ready to go to the moon. This tough little rover, which is about the size of a shoebox, had to drive over obstacles in a thermal vacuum, survive electromagnetic interference, and operate in simulated harsh lunar light. Now that it’s proven itself, the rover will be sent to drive around on and collect data from the moon’s South Pole as part of Astrobotic’s Griffin Mission One later this year.

• Speaking of robots, Integrated Industrial Technologies Inc. out in Blawnox built a robotic tire destacker. It does pretty much exactly what you would expect a robotic tire destacker to do, but still pretty impressive to watch in action.

• In case you missed it, shared a guest post about a new Pittsburgh-focused AI MCP server that the folks from AI@Work (myself included) built as part of a hackathon. If you’ve been wondering what MCP is and what it can be used for, or just looking for a new MCP server to experiment with, this one’s for you!

• SCIFs are finally coming to Pittsburgh. For those of you who are in the security/government space, you know exactly what I just said. For everyone else, though, SCIFs (Secure Compartmentalized Information Facilities) are the industry standard physical workspace for doing work that requires a Secret or Top Secret clearance. So while other, bigger businesses and local branches of government entities may have already had these spaces, this is a huge deal for small to mid-sized local businesses in this segment of tech.

• Canonsburg-based Ansys has gotten approval for its acquisition by Synopsys from every necessary regulatory body except China. Could the ongoing trade tensions between the U.S. and China be affecting this? It’s unclear as of right now, but I’ll be keeping an eye on it.

• Duolingo does Chess now. Is Chess a language? Post your take in the comments.

• For some good local YouTube content to add to your queue, local open source advocate Chad Whitacre posted episode 2 of his Open Path video series. This episode focuses on financial accounting for open source and features a few clips from PyCon, the Community College of Beaver County, and Kennywood. Technical.ly also published a great profile of local startup Sensible Photonics that highlights their electric grid blackout prevention technology and their growth journey.

• Last but not least, warehouse automation company Lucas Systems announced new software that gives real-time recommendations for building pallets. The software performs thousands of calculations per minute across a given warehouse to optimize the way things are stacked on pallets, considering product size, weight, stacking rules, shipping regulations, worker travel distance, safety, route order, and priorities to determine the best way to pick, group, sequence, and stack items. They also have a fun little .gif that visualizes this, which I’ve included below!

  

For the Days Ahead

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• JuliaCon, the Julia programming language conference, will be in Pittsburgh July 21–26th. Tickets are on sale here.

• Women in Product will be taking place at City Kitchen Thursday, August 7th at 6pm. It’s a casual and welcoming space to connect with fellow product managers, share experiences, and build a stronger local product community. RSVP here!

• Did you completely miss the U.S. Open because you just couldn’t carve out the time for it in your schedule? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

Thanks for reading, and have a great rest of your July!

—Austin

Flashback: It’s the beginning of 2025, and you’re just getting back into it after the holidays. Slowly but surely, you fall back into your routines again, start reading your newsletters, and paying attention to what people are talking about on LinkedIn (well…some of it at least). And before long, you’ve heard people mention “MCP” one too many times: “Model Context Protocol.”

You do some research and find plenty of insightful articles talking about how MCP is going to be the future standard for how AI agents will interact with tools—no need to custom build and maintain tools for your agent. Soon, everything will just have an MCP server that you can connect your agent to, and it will “just work.” Sounds great! But how does something that “just works” actually work?

If you’re like the folks at AI @ Work, a Pittsburgh-local collective of AI builders, you can only talk about something for so long before you go and do something about it. So that’s what we did—as MCP hype reached a peak in late March, we set out to run a mini hackathon to learn what really was involved in building and using an MCP server. But we also wanted it to be something that would be useful to the broader community—something that, if we built well, could become a part of how you operate in your day-to-day life.

Building an MCP Server for Pittsburgh Tech/Startup Events

We settled on tackling a simple concept that can quickly become frustrating—helping people navigate all the different tech and startup events happening around Pittsburgh. We felt this was a doable topic that still would be valuable to many people in this community. Our approach was to make an MCP wrapper around a database that contained as many Pittsburgh tech-related events as we could find. This could then be plugged into an agent, giving you the ability to quickly learn more about what events you could attend and even add your own if you wanted.

Finding the Events and Populating the Database

We didn’t need nor want to start from scratch here. There are many great resources out there that list events. Our goal was to make something that consolidates them, allows an agent to interact with them, and brings more awareness to those existing resources. Speaking of these resources, shout outs to these great publications that we used in building this:

• Charles Mansfield's calendar at InnovationPGH

• Technical.ly’s Pittsburgh Calendar

• This blog!

We wanted to make sure our events were kept up to date, so we used two low-code tools—one to scrape websites (Skyvern), and another to orchestrate triggers and populate our database (Make). The scraped event info was then transformed into a standardized format in AirTable for MCP use.

Wrapping an MCP Server Around the Events Database

Simply aggregating the events wasn’t enough for us. We wanted to test out MCP’s ability to make this kind of data universally accessible by an agent so you could do whatever you wanted with those capabilities. Over a few evenings, we spun up a small, but fully functional “events MCP” that allows any agent to search and create Pittsburgh tech events in our AirTable db.

What We Actually Built

MCP Server

We forked Cloudflare’s remote‑mcp‑server template (TypeScript + Hono) and deployed it with Wrangler to Cloudflare Workers. Cloudflare was the cheapest host that natively supports Server‑Sent Events (SSE), which MCP clients need for streaming tool calls (at least, they used to—now it seems to have already moved to Streamable HTTP?)

Tools (API endpoints)

Inside the Worker we exposed two thin, purpose‑built tools:
search_events(start_date, end_date, location, free) → returns matching rows from Airtable
create_event({...}) → inserts a new row
We started with Composio’s generic get‑airtable‑records helpers, but switching to these more narrow tools gave the LLM fewer chances to get the schema or permissions wrong.

Local Dev & Testing

Claude Desktop talks STDIO, so during development we just ran the MCP server locally with “npm run dev”; for remote demos we pointed Claude at our Cloudflare endpoint via npx mcp-remote https://…/sse. A couple of us also experimented with a Python prototype and the community mcp‑gateway, but the TypeScript path had the least friction.

Key Takeaways

• MCP works, but the tooling is young. We hit wrinkles with Claude ignoring an MCP that wasn’t started before itself, Cloudflare env‑var quirks, and Durable‑Object quotas.

• LLMs are surprisingly good at wiring prompts to tool args. We never had to add a separate “date calculator” tool; Claude inferred ISO dates from “next two weeks” just fine.

• Granular, task‑specific tools beat raw CRUD. Limiting the surface area (just “search” and “create”) avoided accidental writes and kept the agent’s reasoning predictable.

• SSE hosting matters. Railway + Docker worked but cost more; Cloudflare’s $5 Worker plan gave us a public URL with streaming out of the box.

End Result

Anyone can drop the MCP URL into Claude or other agent builders, ask “Show me free AI meetups next month” or “Add a Women in Tech happy‑hour on June 5,” and the agent should handle the rest—no API keys, no manual Airtable wrangling required. (By the way, you bet I just copy/pasted our entire hackathon dev Slack conversation into ChatGPT to help write this technical section.)

Putting It into Action

With the MCP now live, we wanted to show some examples of how it could be used, so we made a simple n8n agent that had access to not just this new MCP, but also a personal calendar and a web scraper tool. Here’s where agents get fun. With the right data and context, you can offload the manual task of “finding the right things to attend” to AI instead.

We used a simple chat UI right in n8n to interact with it. It’s powered by OpenAI’s GPT-4.1 model, and has memory so you can chat back and forth with it. But even a simple agent like this shows the potential in well-designed MCP:

Luckily, taking out the garbage isn’t a show stopper (usually), so we can go to PyCon! And this is just scratching the surface—you can set up automations to compare your calendar and the MCP each week and summarize the events you can attend, but filter it down to just events on a specific topic. Ask the agent to create a plan for events based on location, and it will surface only events that are <10 miles from where you’ll be each day. So go ahead, give it your current location and ask it if anything’s happening nearby that you should go to! The world (burgh) is your oyster (pierogi)!

But that’s not all. Since part of what makes MCP powerful is multiple tools, we can also use this same MCP to add events to the PGH Tech calendar so anyone can benefit. Here, I simply asked it to go look at Charles’ calendar and add anything that wasn’t there. The agent used all the necessary tools within the MCP to do it perfectly.

How Does MCP Hold Up? And Should You Dive Deeper into It?

So what did we learn about MCP? Well, it’s got promise. True standards are powerful, but it’s difficult for anything to truly become the standard. MCP has potential and momentum. We showed that it can work and be built, and the benefits of it when it’s all stood up. But we did hit a few snags in getting it up and running, and in the most desperate hackathon hours, we were sometimes left wondering why an MCP was really any better than a simple tool built around an API. There are already multiple tools in it, but if/when we add more concepts to this Pittsburgh Tech MCP (like a job opening aggregator), having one place where all these concepts live as tools becomes significantly more valuable. You only need to add it to your agent the first time around and it updates with more features automatically. There’s also the fact that an MCP should be able to tell your agent how to use it and stay updated. This allows you to avoid manually defining endpoints and parameters in your API tool, then having to remember to keep them updated as well.

Bottom line: Developing a better understanding of MCP is absolutely worth it. Even if it doesn’t take off as the standard, something like it will. And the concepts that make it potentially valuable can only make your own AI and agents better as well.

Want to give the Pittsburgh Tech MCP a try? Email team@aiatworkpgh.com for beta access. And don’t hesitate to reach out to the AI @ Work group with questions or other ideas. For more information, visit http://www.aiatworkpgh.com/.

Lastly, a huge thank you and shout out to all who were involved in this project: April Lutheran, Austin Orth, Benjamin Kostenbader, Dan Gorrepati, Handerson Gomes, Ian Cook, Luisa Rueda, Matt Nicosia, Michelle Powell, and Rex Harris.

Good evening!

So glad you could make it. How was work? Long? Well then, come have a seat on my cyberpunk Lego couch and unwind as I relate to you the various happenings of the Pittsburgh tech scene from the past month.

Let’s get into it.

What Happened This Past Month

The Big Story

Duolingo Goes AI-first, Walks It Back Two Weeks Later

“To teach well, we need a massive amount of content,” Duolingo CEO Luis von Ahn boldly claimed early last month in a memo to his employees. “And doing that manually doesn’t scale… Without AI, it would take us decades to scale our content to more learners. We owe it to our learners to get them this content ASAP.” He then goes on to announce that the company will be replacing contractors with AI and evaluating employee performance based on their AI usage.

This memo was then shared on the company’s social accounts, triggering a firestorm of overwhelmingly negative reactions. The company’s social media accounts were swarmed with frustrated users airing their grievances in the comments section. They argued that Duolingo has already been using AI for its content, and that the new AI-generated content is demonstrably bad. A prominent YouTuber who is also a longtime Duolingo user posted a video decrying the deteriorating quality of lessons on the app, racking up over 500k views. Duolingo’s social channels then went silent for a time, and when they finally started posting again, it didn’t land well (great analysis of this over at Inc.). Prominent business news outlets have since picked up the story, including Fast Company and Financial Times (paywalled).

About two weeks later, von Ahn walked it back. “To be clear: I do not see AI as replacing what our employees do (we are in fact continuing to hire at the same speed as before). I see it as a tool to accelerate what we do, at the same or better level of quality.” It will take some time for the company to recover, though, with its stock down 2% in the past month and their user base still disenchanted.

Note: I reached out to a few folks at Duolingo to ask for their perspective, but none replied as of the writing of this article. A friend told me they have signed an NDA that expressly forbids them from speaking about matters like this.

The Headlines

• Other Pittsburgh locals are having much more fun with AI! Researchers at CMU developed BrickGPT (formerly LegoGPT until they had to legally change the name), an AI model that generates “Physically Stable and Buildable LEGO® Designs from Text.” My favorite is the couch pictured at the top of this post because it’s just hella weird, but you can check out GIFs of the other designs at the link. Local AI builder Rex Harris also launched LipFlix, an AI tool that allows you to make videos of yourself lip-syncing to a song in various visual styles.

• Meanwhile, the Commonwealth has also been throwing its weight behind AI. These efforts include arguing for states’ rights to pass AI legislation (since the budget bill at the federal level currently includes a 10-year moratorium on state-level legislation). They also want to establish a “Commonwealth Opportunity Zone”, which would create a “regulatory sandbox” in which AI companies can start up.

• European company Ardent opened a new data center in Robinson. I’m pretty sure they’re just branding it as an “AI Data Center” because, well, that’s what’s hot right now. But they also say it’s going to have a lot more power and cooling than most data centers because of its AI focus.

• That’s enough about AI—let’s talk food tech. A new local startup Clean Plate Innovations is developing a system using machine learning and computer vision (which… is still AI, but whatever) to generate “real-time analytics on food waste, helping institutions reduce costs, improve sustainability, and make data-driven decisions about menu planning and operations.”

• Heinz launched its Heinz Verified webpage, which maps restaurants that have been verified to serve Heinz ketchup. Users can nominate their favorite restaurant to become verified, and if verified the restaurant will receive “access to insights and consumer trends, free product samples, promotions, branded merchandise, digital resources and complimentary marketing support designed to drive customer traffic.”

• Altoona-based but locally popular gas station/convenience store chain Sheetz was named in the top 300 most innovative companies by Fortune. Between their app and all their self-service kiosk action, I would honestly argue Sheetz is at least 50% tech company, which is why I’m talking about it here. I was recently in a Sheetz, and the auto-grinding single-serve coffee brewers now allow you to make iced coffee as well as regular, which is kinda nuts.

• The local tech job market continues to shrink, a reflection of national trends in the tech sector. The mayoral candidates weighed in on this issue in an interview, with Pittsburgh’s almost certain future mayor Corey O’Connor saying, “To expand equitable access to high-paying tech jobs, the city must build a real partnership with Pittsburgh Public Schools.”

• Remember how the co-founder of Aurora stepped down after their launch for a new role at “an iconic global company?” Well, we now know where he went: General Motors. Is General Motors actually iconic? I’m not sure. Post your take in the comments. Also, since he left, one of Aurora’s partners made them put drivers back in their trucks. That has to feel bad.

• For all the train guys out there, Wabtec showed off some more cool new train tech recently. A robotic, autonomous sled that can slip around rail yards and collect data on train cars called RailGhost and a train remote control system called Pathfinder were two new products that stuck out.

• Looking to take some summer courses? CCAC just launched their new mechatronics lab. What is mechatronics, you ask? Well, it’s “the study of mechanical, electrical and computer systems used to operate and automate modern manufacturing machines.” The program is low-cost and has a graduation and job placement rate of over 90% according to Starr.

• In space news, while Astrobotic broke another rocket, more local funding for space-related initiatives is on the way from a new coalition of local leaders. No one was hurt by the falling rocket, thankfully.

• Finally, for a good evening read, Technical.ly published a great piece highlighting the stories of local Ukrainian tech workers.

For the Days Ahead

• PSN put together an updated list of Pittsburgh’s top coworking spaces. Give one a try! I was recently at CoHatch in Shadyside and it was super chill.

• AI@Work continues to be my absolute favorite local recurring tech event, and we got another one coming up tomorrow, June 12th. If you’re an AI builder and interested in joining, sign up here.

• Looking for a weekly hangout with other folks who are interested in technology and entrepreneurship? Technically Coffee is a great one, 8am every Friday at Bakery Square. Sign up here.

• PyData Pittsburgh’s next talk is on Thursday, June 19th. Jay Palat, from mpathic, will be talking about how machine learning is used to process conversational data. Sign up here.

• Code & Supply is hosting another Build Night on June 30th. Build Night is an impromptu co-working space where you can work on a project, pair, learn, and mingle. Sign up here.

• JuliaCon, the Julia programming language conference, will be in Pittsburgh July 21–26th. Tickets are on sale here.

• Are you having trouble making time to get sunshine this summer or attend any of these sweet events? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

That’s all for now. See you next time!

—Austin

Good morning!

I know, I know. I’m late again with the post. Things have been super busy for me lately, so sorry to keep you all waiting, but I figured better late than never!

Let’s get into it.

What Happened This Past Month

The Big Story

Aurora Launches Fully Driverless Commercial Trucking

After 8 years of development and having overcome several regulatory hurdles, Pittsburgh-based driverless trucking company Aurora has finally gotten its fully driverless trucks making round trips between Dallas and Houston, Texas without any humans in the cab. The trucks have completed over 1,200 miles without drivers to-date, and the company plans to expand service to El Paso, Texas and Phoenix, Arizona by the end of the year.

Since this new broke, Aurora cofounder Sterling Anderson has announced his departure from the company, citing a new job offer at “an iconic global company”, which is somewhat vague, but hey, good for him? He was at Aurora for 8 years and felt it was the right time for him to bow out given that “the technology is on the road, the team is in place to scale it, and the momentum [Aurora has] created in the industry is palpable.”

That said, it’s honestly so encouraging to see a Pittsburgh company pull off a product as wildly complex as a driverless truck. So congrats to all involved!

The Headlines

• Pittsburgh businesses are starting to feel the effects of both tariffs and funding cuts at the hands of the federal government. Several local startup founders recently shared the specific ways these changes are affecting them, including delayed investments and supply chain difficulties, while CMU is suing to block National Science Foundation (NSF) funding cuts. Pitt also “saw five of its NSF grants canceled last month, totaling $1.7 million.”

• While other sectors may be feeling a funding pinch, military funding seems to be going strong. Near Earth Autonomy landed a couple of big contracts with the military this past month. The first is for delivering “miniaturized autonomy systems for U.S. Marines.” These systems, code-named “Fireflies,” are able to autonomously deliver supplies to troops in difficult-to-reach areas because of their small size. The second contract is to create a system to enable, well, self-flying Blackhawks. Feels like something straight out of Call of Duty.

• Speaking of military funding, Gecko Robotics announced a new “extended reality” product that captures and combines thousands of images to generate a 3D model of an aircraft. These 3D models allow teams to not only remotely assess aircraft maintenance needs, but also reveal defects that may be obscured from human eyes. Gecko is now working with defense contractor L3Harris to make the product available to various U.S. military customers.

• The founder of local smart baseball tech company Diamond Kinetics, CJ Handron, was featured in the Technical.ly podcast, and it’s a solid watch/listen! He talks about how the company got started, how their baseball bat sensor product has evolved, what’s been up lately, and what’s next for the company.

• Last but not least, NEXTPittsburgh published a great mini profile of local tech community leader Kashif Henderson. Henderson has been promoting digital literacy locally by fervently working to develop educational and internship programs for local students. Thanks to Henderson’s work, students have been able to shadow IT professionals at Pitt, learn the Python programming language, and develop soft skills to help them in future tech careers. Henderson was named NEXTPittsburgh’s CIO of the Year in the “Tech Community Impact” category, and you can check out the other local award winners in the article as well.

For the Days Ahead

• THIS WEEKEND, May 14th-27th, PyCon, the Python programming language conference, is back in Pittsburgh. Register here.

• Also on Wednesday, May 14th, Dr. Rorry Brenner will be presenting on his Perforated Backpropagation deep learning algorithm at Steel City AI Innovators. Sign up here.

• Wednesday, May 21st, AMA Pittsburgh is hosting its monthly Coffee Connections event on at COhatch Shadyside. The topic this month will be “What Every Website Needs in 2025 (and what you’ll regret ignoring)” presented by Benjamin Kostenbader of Dot Foundry. I’ve had the pleasure of hearing Ben present at other events, and he rocks. You should go. Sign up here.

• AI@Work continues to be my absolute favorite local recurring tech event, and we got another one coming up June 12th. At this past month’s event, we discussed the latest developments in AI agents, Model Context Protocol (MCP), and the merits and pitfalls of “vibe coding.” If you’re an AI builder and interested in joining, sign up here!

• JuliaCon, the Julia programming language conference, will be in Pittsburgh July 21–26th. Tickets are on sale here.

• Are you having trouble making time to get out of the house to some of these cool events? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

That’s all for now. Next month, I promise I’ll be trying harder to be punctual with this post, and also trying to drum up some additional summer reading for you all. Also, if you see me at one of these events, feel free to say hello, and would love to hear from you in the comments if you have an event you’d like me to include in a future post. Thanks, and have a great rest of your weekend!

—Austin

Good morning!

I don’t know about all yinz, but I am exhausted. The national news cycle this past month (and this past week) has been rough. Between that and narrowly losing my company’s March Madness bracket challenge, my emotions have been all over the place. So it was refreshing for me to turn my eyes back towards what’s happening locally when building the usual post for you all. Hope this is refreshing for you to read as well, but if you also need a little encouragement from someone, take 15 minutes, close your laptop, and do something life-giving before continuing on!

What Happened This Past Month

The Big Story

San Francisco Continues Stealing Our Startups

Well, after their big month of funding and being highlighted in Fortune, AI healthcare transcription company Abridge is relocating their operations to San Francisco along with AI agent company Solstis, a recent CMU startup competition winner. To be fair, Abridge’s CEO plans to keep the Pittsburgh office open. But did you know they’re just the latest in a long-running trend of tech companies moving to San Francisco from Pittsburgh? Here’s a short list of Pittsburgh Tech Ex-pats in order of year founded. Some (which I’ve dubbed the Full Relocators) fully relocated their main HQs, but others (the Dualists) kept HQs and/or a heavy presence here in town.

Full Relocators

• BodyMedia (1999), A Wearable Health Tech Company

• GiftCards.com (1999), An Online Gift Card Platform

• ModCloth (2002), Online Fashion Retailer

• Wombat Security (2008), Cybersecurity and Phishing Awareness Training

• NoWait (2010), Restaurant Waitlist App

• Anki (2010), Consumer Robotics/AI

• Mach9 (2021), Geospatial Software/AI

Dualists

• Bossa Nova Robotics (2005), Robotics (obviously)

• ForAllSecure (Mayhem Security) (2012), Cybersecurity

• Petuum (2016), AI/Machine Learning

• Aurora Innovation (2017), Self-Driving Trucks

If Pittsburgh hopes to keep companies like these in town, we probably need to develop a better growth environment for them. The guys from PSN contend that “scaling to a multibillion-dollar status usually means relying on capital and talent from beyond the local ecosystem… Regional leaders must take more thoughtful, longer-term, and decisive steps to actively support high-growth companies.”

Still, there are some examples of local companies who shun the idea of moving to “the Valley,” including most notably Duolingo, which just opened a local center for early childhood bilingual education in an effort to show their commitment to the Pittsburgh community. This is also a luxury they can afford, though, as Duolingo had a lot of access to funding during critical periods of its growth despite not making the jump to a San Francisco office.

So what made Duolingo different from these other firms? Well, there are a few different factors at play here, but let’s just say it’s a little easier to make connections with potential investors when your CEO previously invented reCAPTCHA and sold it to Google.

Special thanks to Shawn Rancatore for helping me build this list. Anyone out there a current or former employee at any of these companies? Or do you know another company that made the jump from the Mon Valley to Silicon Valley? I would love to hear from you in the comments!

The Headlines

• Astrobotic is now partnering with Stamper Technology to send an engraved archive of humanity’s achievements to the moon. The archive, known as GLPH (Galactic Library Preserve Humanity), is engraved on a nickel-based surface “impervious to aging and deterioration” and contains some cool artistic and historical stuff. It also contains some really stupid stuff, in my opinion; for instance, blueprints for an AI-powered sustainable food system that will never actually exist, images from Burning Man, and a boring instrumental Coldplay song. But now I’m going to have to imprint this blog post on nickel and send it to the moon in order for future civilizations to know my opinions on this.

• Niche published their 2025 Best Places to Live Rankings, with Aspinwall taking the top spot for neighborhoods in the Pittsburgh area. The business, like many others, is also facing headwinds introduced by the current administration’s cuts to education funding, highlighted by CEO Luke Skurman recently on LinkedIn.

• Speaking of the aforementioned headwinds, folks are worried about grants for startups disappearing this September when the SBIR and STTR programs are set to expire. These programs have historically been super important for local tech startups to gain footing and compete with larger businesses. Local devs are also lamenting the slash and burn approach the administration is taking with various government entities like 18F, the team that built the U.S. Web Design System.

• The Brookings Institution is now projecting that 34.72% of Pittsburgh workers could see “at least half of their work tasks” significantly change because of AI technology. This is slightly above their 31.24% projection for the nation as a whole, but comparatively lower than most other metro areas in their sample size of 35 markets.

• Technical.ly announced their RealLIST Startups 2025, including one I’ve never heard of before, Moss, which is “an agtech startup with a mission to reduce labor-intensive farming using its AI-powered robotic sensor that analyzes crops to provide data-driven insights on farming practices.” Pretty neat!

• Last but not least, smart warehouse tech company Gather AI launched their new computer vision product, MHE vision. The product “integrates with Material Handling Equipment such as forklifts, auto pickers, pallet trucks, etc.” to track the movement of pallets throughout warehouses, “helping businesses identify bottlenecks, streamline workflows, and reduce unnecessary labor.”

For the Days Ahead

• It’s the last couple of weeks of Lent! Check out Code for Good’s super cool interactive fish fry map to get some tasty fish through Easter.

• AI@Work returns TOMORROW, April 10th. Word on the street is we’ll be talking a lot about MCP this time around. If you’re building things with AI and want to be there, request to join here.

• Inclusive coding group Code and Coffee is 1:30-3:30pm this Saturday, April 12th in Oakland. RSVP on Meetup.

• The Pittsburgh B2B Software Meetup is at Alloy26 on April 28th. RSVP on Meetup.

• PyCon, the Python programming language conference, is back in Pittsburgh May 14-27th. Register here.

• JuliaCon, the Julia programming language conference, will also be in Pittsburgh July 21–26th. Tickets are on sale here.

• Having trouble making time to learn that AI tool you’ve been meaning to try? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

And that’s a wrap! A couple asides—if you’re one of the folks out there searching for a new job, my company’s hiring for several fully remote engineering roles. If you’re interested, message me on LinkedIn and we can talk about a referral. Hopefully I’ll have another local tech economics post coming your way soon, too.

Have a great Easter and a peaceful Spring!

—Austin

Hope you’re all having a great week so far. As usual, the Monthly Beat will be coming out on the first Monday of the month, April 7th. However, I wanted to give a quick time-sensitive plug from my friends over at PyData Pittsburgh for an opportunity to get involved with the local tech community.

PyCon, the annual Python Programming Language Confe…

Read more

Sorry for being a couple of days late here. Just started my new remote job at Machinify on Monday, so it’s been a busy past few days. It’s been a tumultuous month of news at the national level, but let’s once again take a look at what’s going on with tech here in town. I wrote a less lengthy big story this month and opted for more headlines, as there were just a lot of interesting things that happened in February. All companies are in Pittsburgh unless otherwise noted.

What Happened this Past Month

The Big Story

Layoffs Continue, Job Openings Remain Low

I really don’t even need to say anything about this, as yinz are definitely already in the know. But the layoffs continue locally, with Astrobotic letting 16 people (6% of its workforce) go and multiple local branches of government agencies being hit by DOGE’s cuts. Hang in there, folks. I don’t think it’s going to get any easier anytime soon.

In the spirit of supporting the tech workers in this community, here are a few links to jobs in Pittsburgh that I was able to find during my research for this post that may have flown under your radar:

• Aurora has several roles open, mostly senior and above.

• First National Bank has a few new tech roles posted in that big fancy new building of theirs.

• Gecko Robotics has a wide variety of open roles posted on-site in Pittsburgh.

• Kef Robotics has 4 on-site tech roles posted, mixed levels, and travel required.

• Lucas Systems has two software engineering roles posted on-site in Wexford.

• Machinify, the company I just joined, is hiring 9 mostly senior and data-oriented roles and is fully remote.

• Skild AI is still looking to hire multiple roles at multiple levels on-site in Pittsburgh.

• Wabtec is hiring a few tech roles on-site in Pittsburgh, mostly IT and networking.

Best of luck to all of you who are currently job searching!

The Headlines

• AI healthcare transcription company Abridge had a huge month, announcing three new partnerships with various healthcare systems and a $250 million round of Series D funding. Their CEO and Founder was also highlighted with a profile piece in Fortune magazine. My main takeaway is that he listened to Fugazi growing up, so he’s probably chill?

• NEXTPittsburgh published a great piece highlighting the work of Tikkun Olam Makers’ local Pittsburgh chapters. The international organization “creates and disseminates affordable solutions to neglected challenges of people living with disabilities, the elderly and the poor.” Members in Pittsburgh created an open source model for 3D-printed prosthetic limbs, which are both cheaper and more adaptable to various tasks.

• Duolingo, meanwhile, briefly killed its mascot, Duo the owl, then brought him back to life in a bid to get users back on the app and continue growing after its insane growth in 2024. Meanwhile, members of the C-suite continue to dump their stocks at an unprecedented rate.

• Let’s talk robots. A Carnegie Mellon researcher has developed a robot that can locate and remove screws from the backs of flat panel TVs, allowing teams to more quickly disassemble them for recycling much more easily.

• Gecko Robotics is now partnered up with NAES, America’s largest independent power operator, to track and identify power grid maintenance needs with their AI and robots.

• GatherAI is switching to US-made drones. They don’t really give many reasons why in the press release, but you can probably guess.

• Technical.ly published a great interview of a local female machinist and her thoughts on how women can succeed in male-dominated fields like robotics.

• Self-driving company Neya Systems announced the launch of their new N-Drive off-road self-driving product. The company is using Unreal Engine (yes, the game engine) to simulate off-road conditions, allowing them to train their self-driving models without physical testing.

• Also in self-driving, Aurora publicly commented that they’re “optimistic” that the Trump administration will allow them to not put out cones when their self-driving trucks pull over. If they don’t, though, ConeBot™️™️™️ is very much still in play.

• You know what’s cooler than self-driving cars and trucks? Self-driving trains. Smart railroad company Wabtec unveiled a self-driving locomotive prototype downtown. The company has also been gobbling up other companies and signing big contracts to send its trains to Africa.

• PublicSource had a couple of great pieces this past month, digging into how AI is affecting local labor unions and highlighting the voices and perspectives of remote workers. I’d recommend keeping up with the whole Pittsburgh At Work series that they’re doing—excellent stuff!

• Speaking of AI, in what may be one of the most cringe moves ever, the Greater Pittsburgh Arts Council has named their upcoming AI chatbot “Artzy.” Really guys? Really?

• That big new First National Bank building is now open. It’s also real fancy. Amenities include touchless entry, Zoom conference rooms, a café, a fitness facility, and an outdoor space.

• Concerning a couple other local banks, tech workers on the Pittsburgh Subreddit recently discussed what it’s like to work in tech at BNY and PNC. Basically everyone agreed that the culture at BNY was “super toxic.” Comparatively, folks generally agreed that at PNC, the culture varies per team and “the only universal is that the pay is decidedly subpar.”

• Last but not least, local game developer Schell Games is expanding on their success with the VR version of Among Us by launching Among Us 3D. So you will no longer have to have a VR system to play the game, and it’s also cross-play, allowing you to play with folks across PC, Meta Quest, Steam VR, PlayStation VR 2, and PICO. Sus.

For the Days Ahead

• It’s Lent, AKA Fish Fry Season! Check out Code for Good’s super cool interactive fish fry map to get some tasty fish through Easter.

• Inclusive coding group Code and Coffee is 1:30-3:30pm this Saturday in Oakland. RSVP on Meetup.

• The Pittsburgh B2B Software Meetup is at Alloy26 on the last day of the March. RSVP on Meetup.

• The February AI@Work event was absolutely fantastic, and it returns on April 10th. At the last event, we had demos of a thoughtfully designed AI writing tutor, a complex AI pipeline, and a genuine homebrewed deep research agent, as well as tons of great discussion around the latest developments in the space. If you’re building things with AI and want to be at the next one, request to join here.

• PyCon, the Python programming language conference, is back in Pittsburgh May 14-27th. Register here (I will most likely be going!)

• JuliaCon, the Julia programming language conference, will also be in Pittsburgh July 21–26th. Tickets go on sale March 15th here. (I will most likely not be going, but bless you if you’re a Julia user!)

• Trying to make time to go to all these cool events? Check out Confirmed, a locally owned and developed time manager/scheduler app. Plans start at $5/month for a single user, and you can get your first month free with offer code techbeat!

• As always, I’m looking for guest writers! Do you have a local Pittsburgh tech topic you’d like to wax poetic about? Hit me up! Spoiler alert: It’s not a paid gig, but I would be more than willing to host or repost your piece if you want to publish it somewhere else as well.

That’s all for this month! Maybe I’ll see you at Beerhive for some March Madness games between now and April?

—Austin