Security should keep up with how fast we ship
Orvanta Labs is building the autonomous AI security engineer every team wishes they could hire.
Software is written faster than ever — by humans and, increasingly, by AI. Security review has not kept pace. Traditional scanners produce long lists of alerts that teams triage manually, if at all. The gap between "we found a vulnerability" and "we fixed it" is where breaches live.
Our thesis
The same reasoning models that can write code can read it critically, understand how it fails, and repair it. Orvanta puts that capability behind a workflow developers already trust: scan, review, approve, merge. Not a wall of alerts — an engineer that does the work and shows you why.
What we're building toward
A security engineer that lives in your pipeline: understands every service, catches the class of bug that matters before it merges, drafts the fix, and explains the tradeoffs — so shipping secure software is the default, not a separate project.
The team
We're a small, senior team of application-security engineers and ML researchers who have shipped security tooling at fintech and infrastructure companies. We're based in New York and hiring.