I designed and validated a software-engineering benchmark for Project Dynamo to evaluate AI agent behavior, covering dependency graphs, concurrency, and edge-case execution. My verified benchmark was accepted into the project pipeline and merged into the main repository.
I'm building a personal RAG AI assistant that processes PDF, DOCX, and TXT documents with ChromaDB. I evaluate retrieval relevance, coherence, and factual accuracy while refining prompts and semantic retrieval to reduce hallucinations.
As a Software Engineer Intern at ITJobxs, I developed and verified 15+ REST API endpoints and improved system efficiency by 25% through root-cause analysis. I also bring strong competitive-programming problem solving, including 500+ DSA problems and a 1800+ peak LeetCode rating.
