NLP Scientist — Claim Accuracy and Compliance

Ability to work onsite in South San Francisco as needed is preferred. Open to remote candidates able to overlap work hours with PST.

About the role

Develop evidence-grounded NLP systems that verify generated claims against approved evidence sources. Build retrieval pipelines to identify relevant evidence from approved claims, product labeling, clinical studies, publications, and technical references. Decompose compound statements into independently verifiable assertions and evaluate evidence support for each claim. Implement entailment and claim-verification workflows that distinguish supported, qualified, contradicted, and insufficiently supported claims. Design abstention, confidence-threshold, and escalation mechanisms for uncertain evidence. Build traceable decision systems that preserve model versions, evidence sets, citations, and reviewer actions. Develop expert-labeled evaluation datasets with annotation guidelines and inter-annotator agreement measurement. Evaluate false-approval and false-rejection rates separately for high-stakes claim verification. Integrate human review workflows for Medical, Legal, and Regulatory use cases.

Requirements

• Bachelor’s or Master’s degree in Computer Science, Computational Linguistics, Natural Language Processing, Artificial Intelligence, or a related technical field.

Skills

• Advanced Python production engineering with modern NLP frameworks.
• Hands-on experience with evidence-grounded NLP, including hybrid retrieval, natural-language inference, entailment, claim decomposition, and evidence attribution.
• Experience evaluating whether generated answers are genuinely supported by cited sources.
• Experience working with scientific, technical, regulatory, legal, or similarly structured source material.
• Expertise in vector and lexical information retrieval.
• Experience with model APIs and production NLP evaluation infrastructure.
• Experience creating expert-labeled datasets, annotation guidelines, and inter-annotator agreement metrics.
• Experience implementing confidence thresholds, abstention, escalation rules, and safe-failure mechanisms.
• Experience building reproducible and auditable ML decision systems.

Preferred Skills

• Experience combining deterministic rules with ML/model-based judgment.
• Knowledge graph experience linking claims, evidence, references, products, or indications.
• Experience in regulated or high-stakes environments such as legal, financial compliance, scientific publishing, or fact-checking.
• Familiarity with clinical/scientific study design, statistical evidence, and citation methodology.
• Pharmaceutical or life-sciences NLP experience.

Pay

Market Rate - Open Bid. For exceptionally talented individuals, hourly W2 rate of $100-$200/hr.

Schedule

Duration: 6 months to start.