Research Engineer, Privacy and Anonymization

About this role

Build privacy-preserving data infrastructure that detects and removes sensitive information from raw data before it enters AI training and processing workflows. You'll own detection systems combining rules, ML classifiers, and LLMs, create evaluation frameworks measuring privacy risk versus data utility, and ensure robustness across diverse data sources and edge cases.

What you'll do

• Design and implement systems to detect PII, quasi-identifiers, credentials, and sensitive information using rules, statistical models, classifiers, and LLM-based approaches
• Develop production anonymization pipelines that transform data before downstream processing, training, evaluation, and synthetic data generation
• Build evaluation frameworks measuring privacy risk, data utility retention, leakage, and adversarial re-identification resilience
• Engineer systems resilient to schema drift, unusual formats, and sensitive data in unexpected fields
• Benchmark and compare detection approaches across recall, precision, latency, cost, and downstream utility metrics
• Collaborate with engineering, research, operations, and customers to translate privacy requirements into technical policies

What they're looking for

• Python production data and ML systems
• Information extraction and named-entity recognition
• Data classification and sensitive content detection
• End-to-end data pipeline architecture
• Privacy techniques (redaction, masking, pseudonymization, anonymization, synthetic data)
• Privacy-enhancing technologies (differential privacy, k-anonymity, format-preserving encryption)
• Low-latency ML inference and high-throughput data processing
• Experimental design and metrics evaluation