At Google, I built a Python and SQL validation pipeline that profiles large business datasets, flags completeness gaps, and enforces deterministic acceptance criteria. It reduced review rework by 28%.
I also implemented PII redaction and data minimization controls, and designed an audit-friendly workflow that records evidence for each anomaly. That workflow cut stakeholder sign-off turnaround time by 35%.
For downstream AI training, I translated complex data findings into written recommendations and developed JSON Schema Validation checks to catch inconsistencies in submitted artifacts.
At Microsoft, I improved Smart Connected Peripherals telemetry data integrity with ETL checks and anomaly detection thresholds. Earlier, as a Software Development Engineer Intern, I contributed backend validation features and scripts for comparing dataset snapshots.
