SYSTEM Cited by 1 source
Zalando Algorithm Gateway¶
Identity¶
Algorithm Gateway is the re-ranking and personalisation sidecar to Search API in Zalando's catalog-search substrate.
"Algorithm Gateway enriches results with user actions data and re-ranked using the rules engine and our personalization and relevance ML models." (Source: sources/2025-12-16-zalando-the-day-our-own-queries-dosed-us-inside-zalando-search.)
Responsibilities¶
- User-actions enrichment. Consume user-action data (recent clicks, saved items, purchase history) to inform the re-ranking signal.
- Rules-engine application. Apply business rules — category- specific boosts, editorial overrides, merchandising rules.
- ML re-ranking. Apply personalisation + relevance machine- learning models to the organic candidate set from Base Search.
Role as a control-plane lever¶
During incidents, Zalando's Search & Browse team reduces search load by "sampling fewer requests into some heavier ML model integrations" — i.e. fewer requests through the Algorithm Gateway's ML-model path, "falling back to simpler ranking where needed." This is a presentation-layer-controlled graceful degradation: user-visible ranking quality drops, but latency and cluster load drop with it.
Seen in¶
- sources/2025-12-16-zalando-the-day-our-own-queries-dosed-us-inside-zalando-search — canonical wiki instance.