ML Platform Engineer — Feature Store Implementation (Feast)

We have 14 ML models in production and feature engineering is duplicated everywhere. Each model team maintains their own pipelines. When one team fixes a data quality issue, nobody else benefits. When a data source changes, three pipelines break simultaneously. We need to centralise features in a feature store. We've decided on Feast on AWS. We need someone to lead the implementation — designing the feature registry, migrating our 200+ features from scattered pipelines, setting up online (Redis) and offline (S3 + Spark) stores, and building the developer experience for our data science team to onboard new features. You'd own the technical design. Our team (3 ML engineers) implements under your direction.