←
Pete Koomen2,005 posts
Pete Koomen avatar

Pete Koomen

@koomen

GP @ycombinator, cofounded @optimizely koomen.dev

◷ Joined May 2008▧ 64 Photos and videos
Max Rumpf avatar
⟳ Pete Koomen retweeted
Max Rumpf@maxrumpf·22h

Introducing SID-2, a new search model built on compute-scaling swarms. Our swarms find better results than Opus and Astra, while being 8x faster and 20x lower cost. By 2030, AI will write an internet's worth of text every 24 hours. Current search pipelines were built on hand-designed rules, adding more won't scale to meet this incoming volume of data. SID-2 is available on our API today. Try it on your most demanding searches.

2:33
Y Combinator avatar
⟳ Pete Koomen retweeted
Y Combinator@ycombinator·Nov 17

Auctor (P25) is building an AI platform for enterprise software implementations. When a large company rolls out software like Salesforce, ServiceNow, or SAP, the hardest part is often not the software itself, but translating what the business needs into something that can actually be configured and deployed. Since YC, @auctor has raised a Series A from Sequoia, and after rebuilding the product from scratch, saw usage jump 5–10x. In this fireside, the Auctor founders join YC’s @koomen to talk about how they found the problem, launched early, and booked hundreds of demos before they really had the product figured out. They also share why they eventually threw away the product they built during the batch and rewrote it from scratch, a risky decision that became one of the most important decisions they made. 03:28 — Finding the Real Problem to Solve 06:46 — The Dog Photo That Booked Hundreds of Demos 10:13 — When Building for Demos Backfires 13:59 — The Rewrite That Increased Usage 5–10x 18:09 — Why Customers Stayed Through the Rebuild 22:17 — Why Enterprises Still Need Salesforce and SAP 26:21 — Building on Platforms, Not Replacing Them

28:41
Jon Noronha avatar
⟳ Pete Koomen retweeted
Jon Noronha@thatsjonsense·Nov 16

Gamma has a unique window into the AI race. Through it, we’re seeing OpenAI pull ahead. We test every new LLM at scale, and now we’re publishing the winners. What we found: Anthropic excels at nuanced understanding, while OpenAI leads in efficiency and vision; Google is still in the race, and open models still lag way behind their benchmarks. Last month, our spend on OpenAI surpassed Anthropic for the first time in years. But their lead is not secure. The landscape is constantly evolving, and with our own models in progress, the next lab to watch might just be Gamma 👀 Here’s the breakdown of our frontier face-off: research.gamma.app/2026/11/m…

Devon ☀️ avatar
⟳ Pete Koomen retweeted
Devon ☀️@devonzuegel·Nov 16

WE DID IT! Esmeralda just secured its final approval, with a unanimous 4-0 vote! Approvals of this magnitude are extraordinarily rare in California. This has been a long, intense journey I am proud of our team's tenacity. Most of all, I'm deeply grateful

Post attachmentPost attachmentPost attachment
Paul Graham avatar
⟳ Pete Koomen retweeted
Paul Graham@paulg·Nov 15

You work your ass off solving really hard problems day after day for decades, and this is what you get: anyone could have done it. As an investor, I can tell you the limiting factor isn't investment. It's founder ability.

Post attachment
Bence Redmond avatar
⟳ Pete Koomen retweeted
Bence Redmond@BenceRedmond·Nov 14

Day 4 / 46 just shipped embedded visuals! Hoplite now uses HTML/CSS/JS to: - explain patterns behind data - show off new UI elements - detail infra architecture - etc etc etc it also just looks pretty :)

Post attachment

Day 3/46 you can now use Paper directly in Hoplite! try it out today - set up takes a couple seconds

Kieran White avatar
⟳ Pete Koomen retweeted
Kieran White@kierwhite·Nov 12

Pumped to finally launch @mantlemachines to the world, an AI-native maintenance provider for America's industrial machines and robots. Building machines and robots isn’t the bottleneck anymore, keeping them running is. Today @GrahamShips and I change that. Live across the US with multiple customers and 250+ technicians. mantlemachines.com

0:52
Pete Koomen avatar
Pete Koomen@koomen·Nov 13

One of the more ambitious teams I'm working with this fall. It turns out batteries are perfect for rail: they're easy to add to existing locomotives, they're much cheaper than diesel, and it takes less time to swap them than it does to change a crew.

Big day for me: @TagessonEmil and I are launching Ailit, swappable batteries for freight trains 🚂🔋 We met at KTH in Stockholm with the same conviction: the future is electric. North American railroads burn ~$15B of diesel a year. But their locomotives already run on electric motors. So we built the world's largest swappable battery packs to power them directly. → Packs ride behind the locomotives railroads already own → Swap in minutes at stops trains already make → Railroads pay per kWh, priced below diesel, nothing upfront We just finished Superwagon, our first full-scale prototype. Next, we roll it out across our energy network. Rail ran on steam. Then on diesel. We're building what comes next. Backed by @ycombinator.

1:56
Joanne Wang avatar
⟳ Pete Koomen retweeted
Joanne Wang@joannewang_·Nov 11

Epsilab is working with frontier labs to build RL environments for long-horizon enterprise work. Today, low-quality data is an industry-wide problem. We founded Epsilab to bring together real-life data and top 1% domain expertise to solve this. Backed by @ycombinator (F26).

1:25
Pete Koomen avatar
Pete Koomen@koomen·Nov 11

I'm sure whoever decided that trucks should beep while reversing had the best of intentions

Thariq avatar
⟳ Pete Koomen retweeted
Thariq@trq212·Nov 10

we need fewer one shot demos and more tools that show how you can collaborate with models to bring your vision to life

A few folks asked to see how I use the AI Assisted Tween & Paint harnesses, so I put together a little video with Claude! Linking below the workbook :)

2:11
Fabian Zbinden avatar
⟳ Pete Koomen retweeted
Fabian Zbinden@Bartonaut55·Nov 10

Launching 0.security (YC F26): the autonomous cybersecurity team that ships fixes, not findings. Software already writes itself. We make it secure itself. ⚡ 0cloud: fixes straight into every PR 🏢 Enterprise: from web apps down to binaries We've found 30+ Linux kernel 0-days, 500+ CVEs & previously unknown vulnerabilities in government code from multiple countries 🇨🇭🇪🇺🇮🇹 Connect your GitHub now on cloud.0.security get $100 free credits and start shipping with confidence 🤝 Built by the 🐐 @Peak__Twilight & me @Bartonaut55

1:31
Doruk avatar
⟳ Pete Koomen retweeted
Doruk@Peak__Twilight·Nov 8

Introducing: 0 - the open-source AI security harness by @0securitylabs (YC F26). It found 30+ Linux kernel 0-days. Then we got backed by @YCombinator > Found vulnerabilities in software running on 3+ billion devices & government code from 🇨🇭🇮🇹🇪🇺 + Wireshark, node-forge & jsonata > Full-stack pentesting: from web apps & APIs down to binaries & kernels > Multi-model, Reverse engineering, web interface + CLI and MCP for agents. > Security Workflows: parallel agents investigate, reproduce vulnerabilities & test fixes Use it with your coding agent: github.com/0sec-labs/0

0:45
Aaron Epstein avatar
⟳ Pete Koomen retweeted
Aaron Epstein@aaron_epstein·Nov 7

I wish there was a way to mark replies as AI slop, would improve the X experience significantly.