Research Engineer, Privacy and Anonymization

This is a fully remote position, open to applicants in United States, +2 more locations.

📋 Description

• Develop systems to identify PII, quasi-identifiers, credentials, and other sensitive data, while designing transformations based on data type and intended use.

• Create and evaluate detection methods that integrate rules, statistical models, classifiers, and techniques based on large language models (LLMs).

• Construct production pipelines that anonymize raw data prior to further processing, training, evaluation, or synthetic data creation.

• Establish assessment frameworks that gauge privacy risks and the utility of retained data, utilizing recall-weighted metrics, leakage tests, and adversarial re-identification attempts.

• Design resilient systems that can accommodate new data sources, schema variations, atypical formats, and sensitive information found in unexpected fields.

• Collaborate with engineering, research, operations, and clients to convert privacy needs into actionable technical policies and safeguards.

• Manage the complete pipeline for safeguarding privacy while preserving the structure and signals essential for AI-agent training.

• Engage in a 2–3 day work trial as part of the application process.

⛳️ Requirements

• Proficient in Python with a solid background in building reliable production data or machine learning systems.

• Familiarity with information extraction, named-entity recognition, classification, or similar techniques for identifying rare or sensitive content.

• Strong experimental skills and the capability to evaluate methods across recall, precision, latency, cost, and the utility of downstream data.

• Comprehensive understanding of redaction, masking, pseudonymization, anonymization, and synthetic data, along with the contexts in which each is suitable.

• Exceptional attention to detail and the ability to analyze subtle leakage paths, edge cases, and adversarial failure scenarios.

• Experience in building end-to-end data processing pipelines without a fully defined roadmap.

• Practical knowledge of privacy-enhancing technologies such as differential privacy, k-anonymity, secure aggregation, or format-preserving encryption (strong candidates may also possess this experience).

• Background in handling sensitive data in sectors like healthcare, finance, or security (strong candidates may also possess this experience).

• Experience in developing low-latency or high-throughput ML inference and data-processing systems (strong candidates may also possess this experience).

• Experience in navigating unstructured problem spaces and taking responsibility from early research through to production deployment (strong candidates may also possess this experience).

• Experience in early-stage startups and strong communication skills for effective collaboration across teams and time zones (strong candidates may also possess this experience).

• Availability to work hours that overlap 70–80% with either the San Francisco or Singapore time zones.

• Full-time availability.

🏝️ Benefits

• Comprehensive coverage for top-tier medical, dental, and vision plans from Blue Shield of CA (for US employees).

• Complimentary lunch and dinner when working in the office (for in-office employees).

• Company-wide holiday break from Christmas Eve to New Year’s Day, in addition to PTO and paid holidays.

• Unlimited access to tokens for ChatGPT, Claude Code, Cursor, and more.

• Equinox membership (for US employees).

• 401k plan (for US employees).

• Commuter benefits (for US employees).

• Assistance with relocation and visa processes for qualified full-time candidates moving to the US or Singapore.