Video wird geladen...
Video konnte nicht geladen werden
We got into Y Combinator 3 months ago with zero revenue and today Fabraix (YC S26) is at $1.04M annualised run rate. Traditional cybersecurity was built for software that behaves within defined boundaries. AI agents can reason, use tools, take actions, and find ways around those boundaries. The cost... show more
45,265 Aufrufe • vor 8 Tagen •via X (Twitter)
43 Kommentare

@ycombinator @FabraixHQ Has been a pleasure having you in the batch!

@ycombinator @FabraixHQ Couldn't have asked for a better group partner 🧡

@xmuhtasham @ycombinator @FabraixHQ The best in the business

@xmuhtasham @ycombinator @FabraixHQ 🐐

@ycombinator @FabraixHQ Congratulations. A severely underserved market and thankful we have teams like yours thinking about security.

@ycombinator @FabraixHQ AI agents really created their own security industry.

@ycombinator @FabraixHQ Excited for the future

@ycombinator @FabraixHQ Right. The old model assumed software could only do what it was coded to do. Agents get their permissions at runtime, from whatever tool you handed them.

@ycombinator @FabraixHQ Best in class by far

@ycombinator @FabraixHQ learned a couple of things from @heybuildbox

@ycombinator @FabraixHQ these guys are different

@ycombinator @FabraixHQ Pumped for you guys. It really felt like you had put your finger on something when you were talking about your enterprise discovery calls back in May

@ycombinator @FabraixHQ Appreciate it bro 🧡

@ycombinator @FabraixHQ Goat

@ycombinator @FabraixHQ what a legend

@ycombinator @FabraixHQ let's go! incredible future for @FabraixHQ

@ycombinator @FabraixHQ Killers 🚀

@ycombinator @FabraixHQ we're in for some very fuzzy security bugs :)

@ycombinator @FabraixHQ it's just the beginning🫡

@ycombinator @FabraixHQ congratulations guys, what was biggest change pre and post YC?

@ycombinator @FabraixHQ Congrats!!

@ycombinator @FabraixHQ appreciate it man 🫡

@ycombinator @FabraixHQ The zero-revenue YC start still feels like the rare version

@ycombinator @FabraixHQ Bullish let’s go

@ycombinator @FabraixHQ If AI agents are automating all knowledge work, theyre also automating the errors. Asymmetry isnt red-team vs agent - its that your red-teaming agent needs testing when it hits edge cases. Who tests the testers?

@ycombinator @FabraixHQ Is this a hack evaluation on website and system or something else?

@ycombinator @FabraixHQ AI agents don’t break the old security model because they’re smarter. They break it because untrusted text can become a tool call. Are permissions your real boundary, or do you still trust the model to behave?

@ycombinator @FabraixHQ Legend. It's been insane witnessing your journey

@ycombinator @FabraixHQ LESSS GOOO

@ycombinator @FabraixHQ Classic software assumes bounded behavior. Agents reason and tool-call their way around those bounds. If you're shipping agents to production without continuous adversarial testing, you're trusting the demo, not the system.

@ycombinator @FabraixHQ Red-teaming maps what an agent can do. Governance decides what it may do. You need both: the test shows the blast radius, scoped permissions and an audit row per action shrink it before production. Finding holes is step one - making them unreachable is the deployment.

@ycombinator @FabraixHQ Impressive speed. Autonomous red‑teamers will be essential as AI agents scale

@ycombinator @FabraixHQ Congrats, wild 3 months. The part most teams underrate: prompt injection isn't the whole threat model. The nastier one is tool-chaining, where every single action is authorized but the sequence isn't. Do your red-teamers optimize for multi-step paths or single-call exploits?

@ycombinator @FabraixHQ thats a wild ramp for 3 months

@ycombinator @FabraixHQ only the beginning

@ycombinator @FabraixHQ This looks dope

@ycombinator @FabraixHQ Is this Tron?

@ycombinator @FabraixHQ LFG, congrats man 🚀

@ycombinator @FabraixHQ Thanks a lot king 👑

@ycombinator @FabraixHQ LFG Congrats!!! Hard work paid off 🔥

@ycombinator @FabraixHQ Thanks bro 🫡

@ycombinator @FabraixHQ Killer team and killer product🤯🤯

@ycombinator @FabraixHQ Only because I learned from you bro 🫡

![🤖 AI agents are taking over everyday tasks — while you stay in control. KITE AI is building the trust layer powered by Avalanche🔺, enabling millions of secure, agent-to-agent transactions with speed and scalability. They’re creating the platform that makes AI agents trustworthy, autonomous, and ready to transform how we live and work. [Partner Content]](https://image.24vids.com/tw-1965881819103998450/media/G0g3FBmXsAAXlLr.jpg)
