AI Security Researcher

Who we are

We're at an inflection point where AI adoption is accelerating faster than security solutions can keep pace. At Repello AI, we're reimagining AI security from the ground up - merging proactive adversarial testing with automated scale to preempt threats before they're exploited.

We've raised $1.2M from top-tier investors including General Catalyst, Pi Ventures, and we're already working with Fortune 500 companies and AI unicorns to secure their GenAI applications.

Our mission? Give enterprises the confidence to deploy AI at scale - knowing every vulnerability has been found and every interaction is protected.

Role

We need an offensive security engineer who thinks like an attacker first. Someone who sees a new AI agent and immediately starts plotting attack vectors. You'll advance our core technology for AI red teaming and adaptive guardrails while working on the cutting edge of agentic AI and MCP integrations.

What you'll do :

• Develop novel attack techniques against LLMs, AI agents, and GenAI applications
• Build automated red teaming tools that evolve with emerging threats
• Research AI-specific vulnerabilities like prompt injections, jailbreaks, and agent manipulation
• Create proof-of-concept exploits for enterprise AI deployments
• Contribute to our open-source security research and tools

Your New Team

Here's what makes this opportunity special : you'll be working alongside the world's best offensive security researchers. Our team comes from top CTF teams globally and has repeatedly proven themselves against industry incumbents.

We've built careers breaking into the world's most secure systems. Now we're applying that same expertise to secure the AI revolution. You'll learn from and collaborate with researchers who've found critical vulnerabilities, published groundbreaking research, and shaped how the industry thinks about AI security.

⚡ What We Want

The Hacker DNA :

Technical Foundation :

AI Curiosity : (We'll help you level up here)

Why Join Us

Ready to Attack AI systems?

If you're the type who stays up late reading about new jailbreaking techniques and can't resist trying to break every AI system you encounter, let's talk.

How to apply :