At DigiMark, I led the architecture and delivery of production AI/ML solutions, including 12–18 production applications and intelligent workflows. I also mentored a team of 8–12 AI/ML engineers and established standards for system design, testing, and production support.
I architected RAG pipelines that processed 250,000+ documents and served 10,000+ daily queries, with citation generation to support grounded, traceable responses. I also designed agentic workflows using LangGraph, LangChain, and MCP to automate 40+ business processes.
At RaftLabs, I developed production AI systems and LLM-powered applications supporting enterprise workflows. I built RAG pipelines handling 100,000+ documents and asynchronous FastAPI services for high-throughput AI endpoints.
Earlier, at Instinctools, I developed AI/ML solutions and RAG pipelines for enterprise workflows, and built Python and FastAPI services for secure integrations. Across my work, I’ve focused on evaluating and monitoring AI systems, strengthening their security, and making their production behavior more reliable.
