JOB-1540

A seed-stage company that predicts consumer preferences also runs a machine-learning ad network that places individually matched ads inside publishers’ email newsletters. It needs a senior engineer to take ownership of the full matching system, from models through serving to the data platform beneath them, with a future path toward building and leading an ML organization.

The work

  • Build candidate retrieval, ranking and reranking that pair ads with individual recipients
  • Operate the embedding generation and vector-lookup systems behind live ad selection
  • Balance click lift, advertiser return, publisher revenue and reader relevance inside one ranking stack
  • Maintain the feature store plus the route that serves predictions, keeping offline training and online inference consistent and pipelines healthy
  • Apply large language models and other foundation models for tagging subscribers, understanding content and modeling users
  • Lead zero-to-one projects that move prototypes live, covering online learning, feedback loops, ETL/ELT and the dbt warehouse layer
  • Create dashboards and scorecards that give leadership live insight into performance

What you bring

  • Roughly five or more years in production ML or data engineering, including at least three on recommendation, ranking or ads systems
  • A shipped recommendation or matching system that you owned across several layers, from retrieval to evaluation
  • Experience with embeddings and real-time vector search (FAISS, Milvus, Pinecone, pgvector or similar)
  • Self-built data pipelines in SQL and dbt
  • A scientist’s habit of forming hypotheses and testing them with experiments, plus an engineer’s habit of shipping

Hybrid in New York, NY. Candidates must be US citizens or permanent residents; no visa sponsorship.