Security Engineer, Code Review and Vulnerability Research (Remote Contract)

Part-time, remote, paid per completed task. A project we are staffing needs engineers who are as comfortable breaking software as building it: people who can open an unfamiliar repository, find what is insecure or broken, fix it properly, and write up why.
Typical week
Roughly fifteen hours, scheduled around whatever else you do. The tasks are self-contained pieces of security-flavoured engineering work whose results feed into the training of AI coding assistants.
What you would actually be doing
- Hunting down and fixing defects, including exploitable ones, in codebases you did not write
- Auditing code for weaknesses, running pen-test style checks, and rating severity
- Shipping small features and refactors in one of: Python, TypeScript, Go, Rust, Java or C++
- Tuning backend code for performance and hardening it against abuse
- Writing short, clear write-ups of what you found and what you changed
The bar
- Genuine hands-on security experience (audits, vulnerability assessment, pen testing)
- Solid general software-engineering skills, backend included; you can refactor legacy code without breaking it
- You know how secure development lifecycles work and what current attacks look like
- You explain technical findings crisply, in English
No previous involvement with AI projects is needed.
Money and terms
Contractor engagement, output-based. The published range for this project runs up to about $100 an hour depending on task type; exact terms are confirmed at onboarding.
Apply via OpenTrain, where independent AI-evaluation and technical contractors manage their profile and projects in one place.

Work arrangement
Yes