Design an ML search system
OpenAIMachine Learning Engineer · Technical Screen
Last updated
This question evaluates system-design and information-retrieval competencies for machine-learning-powered enterprise document search, covering ingestion and indexing pipelines, hybrid lexical and semantic retrieval, feature generation and ranking, multi-tenant access control, and security/compliance.
Design an ML-powered enterprise document search system. Specify requirements (query latency SLA, freshness, multi-tenant access control, data privacy/PII, cost). Describe the indexing pipeline (connectors, parsing/OCR, normalization), feature generation (BM25 signals, embeddings), storage (inverted index plus vector store) and a hybrid retrieval strategy. Detail the ranking stack (learning-to-rank or neural reranker), query understanding (spelling, synonyms, embeddings), and personalization/feedback loops. Explain online/offline evaluation (NDCG/Recall@k, interleaving/A/B), monitoring, and guardrails (sensitive-content filters). Address scalability, caching, result deduplication, multilingual support, and how you enforce ACLs at query time without leaking documents.
Solution walkthrough is for members
Get the full walkthrough with reference code and complexity analysis, plus Kitty Coach at 300 messages a day. Hints, the code runner and community answers stay free.