Design and build a durable storage layer that preserves source disagreement, history, and provenance rather than resolving them away. Ensure data traceability and reproducibility while supporting point-in-time queries for entity-based analysis.
Extract structured facts from long, narrative, and semi-structured documents while maintaining strict traceability to the source text. Validate extracted data for accuracy and consistency, flagging any ambiguous or unsupported information for review.
You will build and maintain scalable retrieval and grounded generation systems, ensuring all AI-generated text is accurately cited and tested. Additionally, you will own the end-to-end lifecycle of models, including packaging, versioning, serving, and telemetry instrumentation.
You will own the evaluation framework for AI systems by constructing ground truth datasets, designing audit methodologies, and measuring model performance. You will provide independent, objective evidence to support release decisions while ensuring system trustworthiness and fairness.
The engineer will design and implement end-to-end entity resolution pipelines, including blocking, scoring, and clustering strategies. They will also build provenance frameworks to ensure all matching decisions are auditable, explainable, and traceable to source data.
