Software Engineer, Data & AI Systems

Role overview

Build data systems that assemble and maintain a reliable picture of power grid infrastructure from diverse, changing sources, including regulatory filings, operator disclosures, news, market signals, and imagery. The role combines continuous ingestion engineering with language model extraction and geospatial verification. Data quality is central: model-generated records must be measured, traceable, reviewable, and reversible before they inform customer decisions.

Responsibilities - Build and operate continuous ingestion pipelines that handle source formats changing over time. - Create, version, and calibrate prompts and extraction processes that convert documents and news into structured records. - Measure extraction performance against ground truth before model outputs enter the dataset. - Add provenance to records, approval controls for production changes, and the ability to retract records produced by a given extraction process. - Use satellite and geospatial imagery to locate and verify physical grid assets. - Own data quality across the pipeline and dataset.

Requirements - Software engineering experience building and operating data systems or ingestion pipelines. - Practical ability to ship and evaluate language model extraction workflows, including measuring accuracy against reference data. - Attention to traceability, review controls, and data correction across production systems. - Ability to investigate failures in systems that consume inconsistent or changing source material. - Experience or willingness to work with geospatial and satellite imagery data.

Application note - Applicants are asked to send a CV and a short account of either a data system they built and what failed in it, or an LLM extraction they shipped and how they assessed its accuracy.