About the job

CodeWall is building the future of cyber security and on a mission to create the world's best AI super hackers. We are VC-backed by some of the best and just closed our $4M pre-seed round, the largest ever for a cyber startup in the UK. Read our manifesto for why we built it, and our research for what it's already found 👉 https://codewall.ai. Come hack the planet with us!

About the role

Hands-on offensive security, from day one. You'll run our maturing offensive solution on real client engagements, hunt 0-days in our own systems and in the wild, and use that expertise to steer the AI that automates offensive security at scale. You'll break real systems for clients across banking, healthcare, and large consumer brands, and help set the bar for what "as good as a human pentester" actually means. As one of the first hires in the security team, you will have the unique opportunity to shape and define the entire function and build out a world-class team.

What you'll do

• Run our offensive solution on real client engagements — scope, execute, triage output, kill false positives, and deliver findings clients can act on.
• Advise clients through deployment and remediation as a trusted security partner; travel to client sites and switch between engagements at a classic pentest cadence.
• Pentest CodeWall's own systems — find the vulns before our harness can, and be our own first customer.
• Hunt for vulnerabilities between engagements, on open-source and in-the-wild targets.
• Steer the harness: define what it should catch, validate and triage its output, and benchmark agent vs. human until the gap closes.
• Transfer knowledge and findings into the forming internal security team.