At Handshake AI, I design AI-agent evaluation tasks and build realistic environments, reference solutions, and adversarial test suites to assess how frontier models handle reasoning, debugging, and secure software repair. I analyze failure modes and validate results with Docker-based automated pipelines.
Previously, I contributed to AI benchmarking at AfterQuery and developed full-stack applications at SEWASTACK and LETSGROWMORE, including apps serving 10k+ users. I’ve also contributed to open-source EDA tools through the Fossee Internship at IIT BOMBAY, and I’ve solved 1200+ LeetCode problems.

