At Scale AI, I build production multi-agent RAG systems with LangGraph, Pinecone, hybrid retrieval, reranking, and RAGAS evaluation. I reduced redundant LLM calls by roughly 30% and pipeline execution time by about 35% through bounded retries, confidence thresholds, tracing, and metrics.
I've architected high-scale fintech, SaaS, IoT, and business intelligence platforms at InvestMates, Intuit, 100xGain, Pramati Technologies, SIT Project, and Webfosys Networks. Across 11 years, I've led engineers and built event-driven microservices, real-time data systems, and full-stack products using React.js, Node.js, FastAPI, AWS, Kubernetes, Kafka, PostgreSQL, and Redis.
