Data Scientist

What you'll do

• Define and measure data quality across safety, structure preservation, semantic coherence, and customer utility.
• Build evaluation corpora and reproducible experiments; analyze performance across relevant data slices and failure modes.
• Use evaluation results to inform model, pipeline, release, and delivery decisions, communicating uncertainty and tradeoffs.

What they ask for

• At least three years of professional experience in applied science, data science, machine learning, quantitative research, or a closely related role.
• Experience designing evaluations or experiments that changed a product, model, release, or operational decision.
• Understanding of sampling, uncertainty, precision, recall, F1, calibration, agreement, class imbalance, distribution shift, and imperfect labels.
• Ability to write Python and SQL and build reproducible technical artifacts.

Nice to have

• Experience evaluating NER, entity resolution, information extraction, document understanding, multimodal, retrieval, or LLM systems.
• Experience in privacy, de-identification, data quality, model risk, safety, human review, adversarial corpora, or multi-stage pipelines.

Tools

Python, SQL

AI focus

Evaluates de-identification and related AI models, prompts, judges, and pipelines; uses AI tools for analysis, corpus development, coding, review, and hypothesis generation, while independently verifying outputs.