This is the product-level half of the recommender. Give it a set of candidate products and it returns the top ones in order, scored for a specific user from their Strava profile, stated preferences, and, the strongest signal by far, what they actually rebuy. The category/brand recommender answers 'what should this person see'; this answers 'given these products, which ones, in what order,' so the two chain together. The piece I'm proudest of is a leakage-safe repurchase signal that was worth about 13 points of hit@1 in backtest.
This is where models leave the notebook and have to survive contact with real users. I build recommendation and ranking systems that actually run in production: trained, backtested behind a quality gate, and served over HTTP. Plus the LLM pipelines around them. Most of the work isn't the model; it's making it reliable, reproducible, and fast enough that a product can lean on it. I don't call myself an 'AI engineer', I just want to understand these things well enough to build with them.
Browse the full collection →