AWS tutorial: querying structured and unstructured data with Amazon Quick
AWS ML Blog · rss · 2026-11-04
AWS ML Blog walks through combining structured and unstructured data in Amazon Quick. Documents live in S3-backed knowledge bases while structured data becomes enriched datasets; a Space unifies both, and a custom chat agent decides whether to query the knowledge base, datasets, or both, returning cited answers with built-in guardrails.
Key steps
- Create an S3 integration and knowledge base (note: ACL settings are immutable after creation)
- Upload a CSV as a dataset and fix auto-detected data types in the Transform view
- Add dataset-level descriptions and custom usage instructions so the agent translates natural-language questions into accurate SQL
Requires an Enterprise subscription to create datasets.
More from Apps
- Google Labs Expands Dreambeans to India, Brewing Daily Stories from Gemini, Gmail, Photos — GoogleLabs · 2026-11-04
- Mintlify shares five ways teams can use its new Base agent workspace — handotdev · 2026-11-04
- Ex-Vercel designer builds personal framework that compiles MDX into living knowledge garden — floguo · 2026-11-04
- Zethos launches: auto-scans 220k emails at 96% accuracy to write your to-do list — AIwithGhotai · 2026-11-04
- Jot 1.4.0 launches open source: capture selections and screenshots into Obsidian with one shortcut — vista8 · 2026-11-04
- Kagi Search opens no-account trial, touting 75,000+ paying members and zero ads — uwukko · 2026-11-04