Feature Platform Engineer

The work

Halcer is hiring a Feature Platform Engineer to unify offline training features and online serving with lineage eval and compliance teams can trace. Supports classical ML and Gen AI retrieval corpora — no invented platform logos on public pages.

You will

• Define feature definitions, point-in-time correctness, and backfill strategies
• Bridge batch feature pipelines with low-latency serving paths and monitor skew
• Partner with data scientists on training-serving parity tests and with eval engineers on dataset versioning
• Establish access controls and audit-friendly lineage metadata

You likely have

• 6+ years in ML platform or data engineering with feature store or equivalent patterns
• Spark or similar batch compute plus streaming ingestion experience
• Understanding of point-in-time joins and leakage risks in model training
• Written runbooks for backfills, rollbacks, and schema migrations