At IBM, I designed and built a Cross-Modal AI Query System for Logistics Operations that lets operations and procurement users query shipment, SLA, contract, and proof-of-delivery data in natural language. I architected its dual-agent workflow to work across SQL, documents, and images, and supported deployment across AWS services.
At IBM, I also built an NL2SQL-driven academic risk platform with agentic RAG pipelines and benchmarked retrieval and generation quality. Previously, I developed RAG and NLP components at Cognizant and built a customer engagement classification system at Dhruv Technology and System Solutions Pvt. Ltd.

