At UPL Ltd, I led a 60+ member AI, Data Science and Engineering organization and defined the enterprise GenAI and Agentic AI strategy on Databricks over Azure. The AI portfolio was projected to deliver US$34.4M in business value for FY 2025–26.
I architected a reusable agent platform for orchestration, RAG, tool and API integration, and guardrails, cutting AI development and deployment time by 60%. I also built MCP servers for SAP, Salesforce and other enterprise sources, and an Agentic RAG platform that automatically answered 70% of regulatory dossier questions at 95% accuracy.
At IBM, I architected enterprise RAG, machine learning and cloud analytics platforms on Databricks over Azure, and led a team of four data scientists as technical architect. A FinOps recommender system I worked on generated US$10M+ in customer cloud cost savings, while an enterprise RAG and analytics platform achieved a 60% higher query resolution rate.
As an independent AI advisor, I advise startup founders on AI strategy, solution architecture and implementation of Agentic AI, LLM and RAG systems. My earlier work includes a recommender system at MediaKind that increased NDCG by 10% and a fraud detection system at Tavant that saved US$300K.

