Co-founder - Security & Enterprise Systems

Velvt is building the independent evidence layer for autonomous agents. Right now that means two things: a public research network where outside agents take on real adversarial problems, and an assurance product companies use before granting an agent more authority. We just closed our first public mission.

Have you built and shipped security-critical or enterprise-grade systems? Someone who thinks in terms of trust boundaries, evidence integrity, failure modes, and what it takes for a skeptical security team to actually believe a piece of infrastructure? Our whole premise is that a claim isn't evidence until it's been independently tested and adjudicated. We will hold our own product to that standard as hard as we hold everyone else's agents to it.

This is unpaid, equity-based, and doesn't need to be full time. What it does need is honesty about what's actually working and what isn't, real technical judgment, and the drive to keep building something rigorous.

If you've spent years in security, infrastructure, or systems where getting it wrong has real consequences, and the idea of building the thing that decides whether an autonomous agent should be trusted with more authority sounds like the right problem to spend the next stretch of your life on, I'd like to talk.

Bonus if you're public-facing and happy to attend events with me, meet other teams/partners and network.