// applied ai · open role
AI Application Security Engineer
Secure the model layer — the AI features that must never fail quietly.
// about the role
unpwnd is a security company building an autonomous purple team — a swarm of AI agents that audits code, proves what it finds, and hands back the fix. The product runs on model calls, prompts, and the data that flows through them. You’ll define what “secure” means for AI features and hold that line, on both sides of the model boundary.
// what you’ll do
- Secure the AI features themselves: the model calls, the prompts, the data they touch.
- Build defenses against prompt injection, jailbreaks, and unsafe model or tool output.
- Define the security bar for AI features and hold it in review.
- Work across the find/fix seam where the model meets untrusted input.
// what we’re looking for
- Depth in LLM application security: injection, exfiltration, output handling, model misuse.
- You’ve secured — or broken — real AI features, not just read the papers.
- You treat the model as one more untrusted boundary, on top of solid app-sec fundamentals.
- You write clearly; a finding only matters if it lands.
// bonus
- OWASP LLM Top 10 or similar contributions.
- An ML or research background.
- Offensive AI / model red-teaming.
// how we hire
A short async intro, a paid work sample on a real problem, then conversations with the team. No whiteboard trivia, no take-home marathons.
Applying · AI Application Security Engineer
Tell us why you.
No résumé, no cover letter. Your GitHub, where to reach you, and a paragraph on why unpwnd.