At SGR Info Systems Pvt Ltd, I built a Kafka-based distributed RAG architecture that parallelized document processing and brought onboarding time to approximately 30 seconds. I also developed an API-driven RAG indexing pipeline that reduced ingestion time from 20 minutes to 6 minutes.
I’ve architected LangGraph workflows for question answering and built RAG evaluation and human-in-the-loop Text-to-SQL platforms. My work includes hybrid retrieval, semantic schema search, and monitoring retrieval and response quality.
