About

I have spent sixteen years making data usable by the systems that need it. That started with retail point-of-sale integration at Cognizant, in Kolkata, Buenos Aires, and San Francisco. It moved through a supply-chain AI product at Noodle.ai, where I built the pipeline from GBs to TBs and the team from two engineers to eight. At Achieve I unified half a million customer profiles into one governed lake on Google Cloud.

At Zendesk I lead AI-readiness work for the Zendesk Data Platform, a petabyte-scale Snowflake environment. The question I care about most: what does a data platform need to look like before an AI agent can be trusted with it? My answers so far are governed context, semantic layers exposed over MCP, and agentic workflows that keep a human at the review gate.

Outside work I build end to end at Axiomic AI, teach on Built from Bits and The Practitioners Pod, and mentor engineers making the move from pipelines to agents. I am a senior-member applicant at IEEE, a technical reviewer for Packt, and a speaker at IEEE AIC, dbt Summit, and the Applied AI Summit.