Alexander Ollman

My (biased) takeaway from Jensen Huang's comments at GTC is that software tools will continue to be valuable. If you're designing chips you will use CAD tools. If your AI agent produces an image and not a CAD asset, it's useless. You can't verify it, you can't modify it. The same holds for most of knowledge work. We give experts a compressed representation of high complexity information so they can verify it and modify it.

5

Happy Friday! 🖼️ A picture is worth a thousand words! ☺️ This is the latest data from our Central Plant AI optimization solution running at California State University-Dominguez Hills Kenny Seeton. Pretty exciting that we've confirmed that it takes our AI/ML/RL platform a total of TEN (10) days, from “scratch, to completely RE-discover the data points needed, classify them, "map" them, build its OWN model, train itself, and INCREASE the Central Plant efficiency by FORTY-EIGHT PERECNT! From 0.81 kW/ton down to 0.39 kW/ton! 🚷 NO human intervention! Below is a snapshot of the CP. Restart 5/5/25. We deleted ALL of the learnings from the last 2 1/2 years. About 2GB. 1 week baseline. 10 days RL. Notable observations: 1. Even though the Plant Efficiency increased by 48%, the Chiller is STILL "holding" the CHW-S temp set-point of 47 DEG F (46.5 DEG F Actual). 2. The one-thousand TON (rated) JCI/York YMC2 Chiller is "putting out" 668 TONS of cooling (67% of its RATED capacity), YET the Chiller itself is only "loaded up" 39.4%! This means the Chillers CAPACITY (output) INCREASED significantly. 3. The original BMS algorithm was programmed to set the CW Temp to Outside Air Wet-Bulb (OA-WB) PLUS five (5) degrees. Which is standard practice, and has been, for decades. However, our AI trained itself and controlled the CW Temp everywhere from +4 degrees above OA-WB to TWENTY-FOUR (+24) degrees above OA-WB. And everywhere in between. Every five (5) minutes. 288 cycles (solutions) per day. The new set-point is computed and sent from the cloud. Takes 4-5 secs. Let's talk about "guardrails". 😉 YES! 😁 They're there. Low: 60.0 DEG F to a High: 90.0 DEG F. In fact, these same "guardrails" are typically already IN PLACE, with your existing BMS logic. AND there's mechanical safeties, ON the Chiller(s) and equipment. We take the safety and security of your organization VERY seriously. But here's the thing about facil.ai: How can you TRUST our AI? Because you can TRUST the people who built it! And install it. Our partners are the same BMS service providers, consultants, mechanical contractors, ESCOs, Property Managers, Peers and friends: Who are serving you right now! Ready to learn more? Contact us: https://lnkd.in/gVutYEiX Want to explore being a partner? https://lnkd.in/gigtx2vx Want to invest in the future of AI Digital Technology? We're putting together a crowdfunding growth round with StartEngine https://lnkd.in/gwcSRR3b Please sign up on the list if interested: https://lnkd.in/gJ_2yCr3 Let's start a conversation! I'm always ready to "talk shop" 🤗 Shoot me a note on LinkedIn or e-mail: keith@facil-ai.com 🦄 We’re thrilled to be in conversations with several large infrastructure companies as partners or acquirers. Inquiries to our interim CTO John Majeski BUILD🛠️GROW📈FUND💰EXIT🦄 ❤️ Our customers love us! 🚀 facil.ai is proud to be a Portola Valley Partners portfolio company. #innovation #technology #management #creativity #future

70

13 Comments

Podonos just raised $2.4 million in pre-seed funding, and the signal is loud and clear: the voice AI industry finally has someone measuring the sound, not just making the noise. Based in Los Gatos, Podonos is building the infrastructure layer that decides whether voice AI feels real, human, and ready for primetime. The industry is racing toward a projected $47.5 billion by 2034, but growth without standards is chaos. Podonos is making sure the hype actually holds up when you press play. At the center is ✦Soohyun Bae, PhD in Computer Engineering from Georgia Institute of Technology and Y Combinator W22 alumnus. His track record stretches from engineering roles at Google Maps to leading AR mapping at Niantic, plus co-founding Bobidi and TickTock AI. He has seen what happens when technology scales without proper #guardrails. With Podonos, he's betting the future of AI isn't just about how smart the models get, but how believable and trustworthy they sound. Podonos evaluates models across naturalness, similarity, emotion, recognition accuracy, pronunciation, tone, and resilience. Add in #personaconsistency, and suddenly you're not just testing machines, you're testing their ability to pass as human. The twist? They deliver results in under twelve hours, when the legacy approach drags on for two months. In AI time, that's the difference between relevance and irrelevance. Investors caught the beat. Serac Ventures, led by Kevin Moore, took the lead, backed by NAVER D2SF, the venture arm of South Korea's NAVER, and KAIST Venture Investment Holdings. Together, they've now backed Podonos twice, with this $2.4 million round stacking on top of a $750,000 raise in March 2025. Total funding sits at $3.15 million, and every dollar is pointed toward scaling #engineering, refining #analysistools, and launching into the U.S., Europe, and Southeast Asia. This isn't theory. Podonos already serves Resemble AI, Play AI, and Sanas AI, hitting six-digit ARR by November 2024 with just four employees. Their secret weapon is reach: 150,000 evaluators spread across nine languages and thirteen locales, forming a human-in-the-loop platform that blends speed, rigor, and global coverage. That reach lets Podonos deliver evaluations at a scale and pace that no competitor has matched. The strategy is ambitious: expand into high-demand sectors like #healthcare, #finance, #gaming, and #advertising, then move beyond voice into #multimodal model evaluation, spanning video AI and large language models. The mission isn't just to check if AI works, it's to decide if AI works for people. #Startups #StartupFunding #EarlyStage #VentureCapital #PreSeed #AI #AIInfrastructure #Voice #VoiceAI #VoiceTech #Infrastructure #Technology #Innovation #TechEcosystem #StartupEcosystem #Hiring #TechHiring If software engineering peace of mind is what you crave, Vention is your zen.

6

1 Comment

The coolest thing about building a robot on Viam is all the things you don't have to learn: inverse kinematics, motor control systems, motion planning algorithms. Viam abstracts away that complexity with a simple API call so you can focus on being a good developer. Our mobile team proved this when they built a robot arm controller using just Flutter and ARKit in just a handful of days, and use it to grip colored blocks. This is what we're building toward at Viam: making robotics development more accessible to software engineers. Read how Clint Purser, Julie Krasnick, and Martha R. Johnston built this robot and everything they learned along the way: https://lnkd.in/eSTSru2R.

62

3 Comments

The end of human drivers? Tried Tesla Robotaxi in the Bay Area for the first time, 8.8 miles from Meta SF to Daly City station for $18. Unlike Waymo, Tesla’s Robotaxi can take the freeway. The ride was smooth, and the decision-making felt far closer to human than “bot-like.” What surprised me most: it’s just a regular Model Y. With the built-in rear screen, any Tesla could potentially be activated as a Robotaxi. That’s scalability at an entirely different level. Waymo might have an early lead, but Tesla’s vision-only approach gives it massive coverage and faster rollout potential. If this scales, Uber, Lyft, and even car ownership itself could be disrupted faster than expected. This is exactly why I quit my job to go all-in on ElasticDash We’re entering a new industrial revolution, the real question is: will you shape it, or be forced to adapt?

15

2 Comments

Another kiwi company punching well beyond its weight …. OpenStar Technologies Will the future generation be Fusion, and can New Zealand become the sandbox for the world? https://lnkd.in/eEdM-CGC

6

I spoke at the Portland AI Engineers meetup last week about a question that keeps many AI engineers up at night: how do you actually know if your AI works at scale? Last year, Cursor's support bot hallucinated a device policy that didn't exist. Multiple users asked about logging in on different devices. The bot confidently claimed they had to log out first to switch devices or pay for a second subscription. That policy never existed. The bot hallucinated it. Some users believed the made-up policy. Others argued it was wrong. Chaos ensued and users started canceling their subscriptions. The CEO had to intervene publicly. The haunting question: how many users got bad answers and silently switched to the competitor? Here's why this keeps happening. Traditional testing can't catch it. You can't write `assert chatbot_response == "correct answer"` when the output is different every time. Unit tests assume determinism. LLMs are stochastic. Generic LLM-as-judge evaluations don't catch it either. A generic eval asks "does this answer the user's question?" The Cursor bot did answer the question. It just invented the answer. What you actually need to check is "is this factually consistent with our latest policies?" That gap between generic and contextual is where AI products break. The solution is encoding domain expertise through labeled examples. Not "does this avoid hallucinations" but "does this match our actual support documentation?" 20 examples showing the judge what correct looks like in your specific context. The attached video shows how to build contextual evals in Truesight. Same LLM-as-judge approach you'd use for generic checks, but grounded in labeled examples from your actual domain. This is the technique that would catch hallucinations like Cursor's before they reach users. Once you have this built, you integrate it everywhere. Development, pre-deployment regression checks, production monitoring, real-time guardrails. The eval becomes the continuous quality signal that keeps you out of the reactive loop. Full talk with the Cursor breakdown and all demos: https://lnkd.in/gwRBtpA8 If your evals aren't catching what actually breaks in production, try Truesight: https://lnkd.in/gfbpP8dA What evaluation challenges are you hitting as you move AI to production?

16

7 Comments