AI Engineer Intern (Contract)
- Developed specialized training datasets and evaluation environments targeting frontier model failure modes, improving model reliability in adversarial and edge-case scenarios.
- Designed and evaluated 1,000+ task instances advancing Claude Sonnet in software engineering workflows specifically Git operations, pull request review, and repository issue resolution at scale.
- Built and tested autonomous AI agents capable of multi-step task execution across complex software engineering scenarios; identified and closed capability gaps in agent reasoning and tool use.






Socials