At Samsung R&D Institute India, I own and operate production data pipelines for Samsung Wallet India, from Java and Spark parsers to Airflow on EMR and real-time Kafka ingestion. I redesigned a parsing pipeline that was producing partial output, cutting its runtime from most of a day to a couple of hours and reducing AWS storage and compute costs.
I also cut promotion-analysis query times from hours to under a minute, shipped Protobuf usage logging in an Android app, and led a Vue and Express feature rolling out to Samsung Wallet users. More recently, I’ve added LLM-generated insights to daily reports and am building a privacy-conscious MCP layer for natural-language analytics queries.

