# Design a Personalized Weekly Deals Service Design a service that ingests deals weekly from several pipelines. Each deal has a start and end time and can be active only within that window. Users browse active deals on a main page, twenty per page by default, apply filters, and receive a ranking informed by browsing history. A user may claim a given deal at most once. Cover ingestion, validation, storage, active-window queries, ranking, pagination, claim correctness, and failure recovery. Do not assume ranking must be a machine-learning model; begin with the simplest approach that meets the stated behavior. ### Clarifying Questions to Ask - Can different pipelines send the same deal or later corrections? - Is claim inventory limited, or is only per-user uniqueness required? - How fresh must browsing history and ranking be? - Must pagination remain stable while ranks or active windows change? - Which filters and sort options are required? ### What a Strong Answer Covers ```premium-lock What a Strong Answer Covers ``` ### Follow-up Questions - How would globally limited inventory change the claim transaction? - How would you explain a recommendation to a user? - How would you backfill corrected deals without disrupting pagination? - What privacy limits apply to browsing-history features?