Principal Data Scientist – Identity and Matching Models

This is a fully remote position, open to applicants in Arizona, +17 more states.

📋 Description

• Develop and enhance models that match records and assess confidence in each outcome.

• Create labeling and measurement strategies, which include evaluation sets and manual verification processes.

• Leverage AI models to facilitate large-scale data labeling and evaluate results against human assessments.

• Deploy models to production on AWS utilizing ML engineering and DevOps practices.

• Oversee retraining timelines, the model registry, and records of model/data lineage.

• Review each data build prior to release, identify modifications and regressions, and communicate findings to product and sales teams.

• Collaborate with the Product Owner to transform insights into actionable stories with clear objectives and prioritize research initiatives.

• Document methodologies, evaluate teammates' code, and engage in the DS/ML working group.

• Mentor an expanding data science team.

⛳️ Requirements

• Over 6 years of hands-on experience in data science.

• A Bachelor's degree in a quantitative discipline is required, or equivalent professional experience.

• A Master's degree is preferred.

• Proven experience in taking a model from concept to production in a cloud environment such as AWS SageMaker, Databricks, MLflow, or similar platforms.

• Skilled in monitoring models and assessing performance following deployment.

• Proficient in working with extensive datasets (100 million rows or more) utilizing Spark or similar tools, Python, and SQL.

• Experienced in evaluating models without a clean answer key, including sampling and manual verification of results.

• Capable of discerning genuine improvements from noise and articulating results, limitations, and trade-offs in straightforward terms.

• Experience in maintaining models as data evolves, including updating outdated labels and models and documenting procedures.

• Knowledgeable in handling personal data responsibly and adhering to opt-out, retention, and lineage policies.

• Familiarity with record matching or deduplication, AI-assisted data labeling, graph methods, or translating raw model scores into actionable confidence is advantageous.

• Must be eligible to work in one of the following US states: AZ, CA, CO, FL, GA, KY, MD, MI, MO, NJ, NV, NC, PA, SC, SD, TX, VA, WA, or WY.

🏝️ Benefits

• Annual bonus program.

• Stock options.

• 401K matching.

• 100% coverage for medical, dental, and vision.

• Unlimited paid time off (PTO).

• Resources for mental health support.

• Reimbursement for home office equipment.

• Fitness reimbursements.

• Financial support for educational courses.