ML Engineer — Recommendation System Rebuild

Our content recommendation system was built 4 years ago and hasn't been meaningfully improved since. Click-through rates are declining and our users are telling us recommendations feel stale. We need an ML engineer to lead a full rebuild — new features, better models (matrix factorisation? two-tower? transformer-based?), and an A/B test framework to validate improvements. You'll work with our existing data team and have access to 3 years of user interaction data.