At Seatrium Digital, I own solution architecture and production deployment for enterprise multi-agent AI applications using AWS Strands SDK and Amazon Bedrock. I designed a four-stage agentic architecture for SQL retrieval, document retrieval, orchestration, and response synthesis, with a focus on secure access and grounded answers.
At HCL Technologies, I developed conversational AI, NLP-to-SQL, and agentic RAG applications, helping reduce process turnaround time from two days to four hours. I also productionized multi-cloud AI deployments and MLOps/LLMOps workflows; that work reduced manual intervention by 75% while maintaining 95%+ CSAT. My projects include a browser-based micro-expression system that achieved sub-100ms inference and a secure enterprise RAG platform that achieved 99% data-access compliance.
