Design a multi-tenant assistant that answers employees' questions from their organization's private documents and cites the passages supporting each answer.
Question
Design a multi-tenant assistant that answers employees' questions from their organization's private documents and cites the passages supporting each answer.
## Requirements
- Cover document ingestion, parsing, chunking, embedding, indexing, retrieval, reranking, generation, and citations.
- Enforce document permissions at retrieval time, including group changes and revoked access.
- Refresh changed documents without serving mixed or stale versions longer than a defined bound.
- Isolate tenants, encrypt sensitive data, and make access auditable.
- Handle empty, conflicting, stale, or low-confidence retrieval without inventing an answer.
- Define latency, scale, availability, and cost assumptions before sizing components.
Describe the external chat and ingestion APIs, metadata stored per chunk, asynchronous jobs, and the online request path.
## Follow-ups
1. Evaluate retrieval independently from answer faithfulness and end-to-end usefulness.
2. Support permission-aware caching and conversational follow-up questions.
3. Add human feedback, incident investigation, and safe reindexing.
4. Compare lexical, dense, and hybrid retrieval.
The public brief is identical for people and crawlers. Account access changes tools, not the indexed description.