Lead a team of 3 data engineers and partner with the COO and analytics leadership on data strategy.
Rebuilt the enterprise data platform on Microsoft Fabric as a medallion lakehouse, joining Opera, Amadeus, Salesforce and Delphi data.
Moved 250+ Azure workloads from personal logins to managed identities, with no major failures.
Cut cloud costs through an infrastructure audit, and built AI data quality checks that run five times a day.
Led a two-person Cloud Analytics team for Nordstrom, building inventory tracking.
Tuned Spark clusters for 5x higher efficiency and a 50% cut in infrastructure costs.
Shipped ML models on Azure Databricks that predict claims.
Delivered analytics and automation for enterprise clients with Microsoft Power Platform and Power BI.
M.S. in Business Analytics, Boston University, Questrom School of Business, 2023
B.Tech., IIT Bombay, 2018
Microsoft Certified: Azure Data Engineer Associate (DP-203)
Forecasts and budgets lived in a vendor tool, an Access app and spreadsheets.
Took a design prototype to production with 2 engineers and AI coding agents, behind automated release checks.
In production in 10 weeks, now in pilot. Daily, monthly and P&L views agree to the penny.
Forecasts drove decisions, but no one had measured how far off they were.
Tested every model against real results, then fixed the defects the numbers showed.
Error on past data fell from 40% to 16%. Monthly forecasts now cover 100+ hotels.
An 8.2 TB production database had no documentation.
Built an AI agent that maps and documents it, and asks a person before any change.
600+ changes made. Every one approved by a person.
Every fix is checked in the report people use, not only in the table.
600+ AI changes so far. None ran on their own.
We retracted our own "model beats humans" result after a units check.
Nothing gets dropped because a tidier story came along.
03 / TALKS
SIM Boston
Build Trust in Your Data with Transparency
Mar 19, 2026
devopsdays Boston
We Gave the Agent Production Credentials
Oct 19-20, 2026 UPCOMING
Oct 19-20, 2026
Cambridge, MA
We Gave the Agent Production Credentials
Conference session
Oct 10, 2026
Pittsburgh, PA
The Best Model Was a Ratio: Hybrid Forecasting on Microsoft Fabric
60-minute session · Data professionals at a free, community-run conference
Oct 2, 2026
Philadelphia, PA
We Gave the Agent Production Credentials
Ignite talk · DevOps and platform engineers
Mar 19, 2026
Boston, MA
Build Trust in Your Data with Transparency
Interactive session · CIOs and technology executives