At DataLyzer Technologies Pvt. Ltd., I built an ML-based process-drift and special-cause detection model for a production SPC analytics product. Tuning its thresholds cut false quality escalations by about 40%.
I also built data ingestion and validation pipelines that fed internal SPC/OEE dashboards, reducing manual data-preparation effort by about 65%. I developed forecasting models for quality and productivity KPIs across the Qualis SPC and Gage Management product lines.
At Bharat Electronics Limited (BEL), I automated XML-to-SQL Server ingestion and SQL-to-Excel reporting workflows, cutting manual reporting time by roughly half. I also built an Excel-driven validation and write-back workflow to reduce data mismatches and reconciliation effort.
In my projects, I built IncidentPilot, an incident investigation agent with read-only tools, a human approval gate, and deterministic evaluation. I also developed clinical research and contract-risk pipelines, and a full-stack intrusion detection system.
