Note: The job is a remote job and is open to candidates in USA. Mind Moves LLC is a women-owned firm helping businesses and governments develop and deploy human-in-the-loop AI solutions, with expertise in machine learning and AI. The Synthetic Data Engineer Specialist will design, build, and validate privacy-preserving synthetic data pipelines for sensitive genomic and health datasets, while supporting AI system prototyping, testing, and red-team activities. The role also involves assessing privacy risks, documenting methodology and data provenance, and collaborating with technical, institutional, and federal stakeholders.
Responsibilities
- Design and build synthetic data generation pipelines (statistical modeling, generative models, or agent-based simulation) that replicate the structure and statistical properties of controlled-access datasets
- Generate realistic synthetic test/dev datasets (e.g., genomic sequence-style data modeled on dbGaP) to support prototyping, MVP development, and red-team testing without exposing real participant-level data
- Apply and tune privacy-preserving techniques, including differential privacy noise injection, to bound re-identification and membership-inference risk
- Validate synthetic data fidelity — statistical distribution matching, utility-preservation testing, and downstream model performance parity against real data benchmarks
- Assess and document residual privacy risk (e.g., re-identification, linkage, membership inference) for each synthetic dataset release
- Collaborate with the Red-Team/Adversarial Security Lead and Biomedical Privacy & Gateway Engineer to supply adversarial and edge-case synthetic scenarios for testing
- Contribute methodology documentation and data provenance records for review by institutional Data Access Committees (DACs)
- Collaborate with cross-functional teams across federal agencies and supporting partners
Skills
- Demonstrated experience building synthetic data generation pipelines using statistical modeling, generative models (e.g., GANs, VAEs, diffusion), or agent-based simulation
- Experience engineering and transforming large-scale genomic formats (e.g., multi-sample VCFs, PLINK binary sets, Hail MatrixTables) alongside complex, longitudinal clinical and phenotypic data
- Advanced proficiency in Python and SQL, with hands-on experience using synthetic data, privacy, and scientific libraries (e.g., Opacus, Diffprivlib, PyTorch, Hail, Polars, or Spark)
- Data engineering and ETL skills — pipeline design, schema modeling, API integration
- Experience taking synthetic data pipelines from proof of concept through to production-grade, reusable tooling
- Strong communication skills; ability to document methodology and risk assessments for non-technical and institutional stakeholders
- U.S. work authorization for 1099 contracting; Mind Moves cannot offer sponsorship for this contract
- Contract hours: 20-25 hours/week
- Knowledge of AI/data governance frameworks such as NIST AI RMF and federal (NIH/HHS) compliance requirements preferred
- Background or experience in biomedical, bioinformatics, or life sciences preferred
- Work locale: Remote in the US (EST time zone preferred)
Benefits
- Remote work in the US
- Gain hands-on experience with privacy-preserving data generation for frontier AI/ML pipelines
- Operate at the intersection of data engineering, privacy, and federal data governance in a creative, mission-driven environment
Company Overview