9. Design Search Autocomplete
Return useful prefix suggestions quickly while refreshing rankings and removing unsafe entries.
Start with a template. Work through each step. Ask Coach when you need a second opinion.
Company tags are community-reported. Counts on cards show how many people reported that design.
Return useful prefix suggestions quickly while refreshing rankings and removing unsafe entries.
Handle document changes, ACL revocation and deletion across a retrieval pipeline.
Hold seats with one conditional write, pay by authorize and capture, and put ten million fans in a fair waiting room without selling a seat twice.
Search 10 million businesses by place, words and category, keep every average rating exact, and keep fake reviews out.
Search a billion posts a day by keyword, newest or most liked first, with an inverted index you build yourself.
Stacks of nearby people in under 300 ms, 2 billion swipes a day, and a match never missed even when both like at once.
Search a billion documents in 200 ms: inverted indexes, analysers, Lucene segments, shards and replicas, query then fetch with BM25, fed by CDC.
Radius, k-nearest and map-box search over places that rarely move and drivers that ping every few seconds: R-trees, geohash, H3 and S2 cells, Redis GEO and OpenSearch. One-hour boards for junior, senior and staff, with the theory.
Nearest-neighbour search over a billion embeddings: HNSW, IVF-PQ and disk graphs, filters, segments, sharding, and hybrid search with re-ranking.
Design a team chat app like Slack: channels and DMs in real time, threads, unread counts and search. One-hour boards for junior, senior and staff, with the theory behind them.
Ten terabytes of logs an hour from host agents through Kafka into tiered OpenSearch and S3, with regex search, live tail and exceptions grouped into issues.
One set of tags across Jira issues, Confluence pages and Bitbucket pull requests: batch renders, tag pages that never leak, suggestions, popular tags.
Embed a billion chunks on GPUs, keep a k-NN index fresh through CDC with versioned writes and provable deletes, and migrate models blue-green.
Find the right videos for a typed query among billions: BM25 and a dual encoder over text, speech and frames, LambdaMART on debiased clicks, human raters.