Data Engineer

Sunset•New York, NY

6d•$180,000 - $280,000•Onsite

About The Position

Sunset is building the data layer for real-world AI training. We work with frontier labs to turn messy, multi-modal enterprise data into the highest-quality training data on the market — sourced from the hundreds of venture-backed startups we've helped wind down. We're a fast-growing team based in-person in Dumbo, Brooklyn. Backed by Floodgate, Afore Capital, Hustle Fund, and incredible entrepreneurs. As a Data Engineer at Sunset, you'll own the pipeline that turns raw, chaotic enterprise data into the highest-quality training data on the market. One of our core technical problems is entity resolution and de-identification across different sources and modalities. The deeper challenege is understanding the node structures and linkages well enough to effectively reconstruct the business world this data comes from.

Requirements

You are a product minded engineer and have shipped data pipelines at scale
You have strong Python and are comfortable across NER, record linkage, and coreference
You want to own a hard, ambiguous problem end-to-end rather than wait for a PRD
AI is deeply integrated into your workflow and life

Responsibilities

Extending our entity resolution pipeline to handle new modalities — think audio transcripts, design files, or embedded references inside PDF contracts
Building coreference resolution across Slack threads, email chains, and Linear comments so that "me," "him," and first-name mentions all resolve to the right canonical entity
Designing the de-identification layer that replaces PII with stable pseudonyms while preserving every relationship across every source
Figuring out how to ingest formats we've never seen before, fast