At Optum, I engineered PySpark and AWS Glue pipelines for healthcare claims, cutting recurring batch runtime by 32%. I also modeled claims and eligibility marts in Snowflake so operations teams could run recurring reports faster.
I built reusable claim-level feature tables to support production fraud and overpayment experiments. I also developed a Python and Amazon Bedrock PR assistant that posts cited review comments in GitHub Actions, reducing manual database code review effort.
At Bajaj Finserv, I designed Azure Databricks and Data Factory pipelines for consumer-loan and repayment data, processing more than 20 million records. I also piloted a GPT-4 RAG search service that helped analysts find lending-policy information faster.
Earlier, as a SQL Developer Intern at Tech Mahindra, I assisted with data loads and query tuning, and built Power BI datasets for sales operations. I hold an M.S. in Data Science from the University of Maryland.

