- Assess enterprise data quality across connectors, including topic coherence, domain depth, completeness, and consistency.
- Design and automate pipeline validation suites for schema checks, completeness, drift detection, and reconciliation across raw, processed, and sanitized stages.
- Design, train, and evaluate ML models that detect sensitive entities in text and image-based documents.
- Build replacement pipelines that map the same entity consistently across files in a corpus.
- Build adversarial de-identification test sets covering edge cases, obfuscated identifiers, multilingual entities, OCR noise, and formats designed to evade detectors.
- Measure and report de-identification performance by data class, including precision, recall, leak rates, false-negative audits, and replacement consistency.
- Implement CI/CD regression gates for data quality and sensitive-data checks.
- Run sampling-based human-in-the-loop audits and maintain audit trails as compliance evidence.
- Partner with engineering on root-cause analysis of inconsistencies or leaks and drive fixes to closure.
PythonSQLMachine Learning