At Atlas AI Labs, I deliver agentic RAG assistants for enterprise knowledge bases, taking client GenAI use cases from prototype to deployment. I added source citations and automated evaluation guardrails so answers can be traced to source documents.
I also built and deployed Atlas, an autonomous agent platform that researches the web, runs Python in a sandbox, and answers from documents; it reached 96% task success in a live evaluation. In CRAG – Corrective RAG, Ask Any PDF, I built a self-correcting retrieval system that improved faithfulness by 88% over basic RAG on a RAGAS benchmark.

