Bring structured and unstructured data together with Amazon Quick
aws ·
This article explains how Amazon Quick enables organizations to unify structured and unstructured data through a single conversational AI interface. The solution architecture consists of three layers: a data layer that indexes S3 documents in knowledge bases and loads CSVs as enriched datasets, a workspace layer called Spaces that combines these sources, and an agent layer that queries them simultaneously. Users can create custom chat agents with specific personas and communication styles to analyze business metrics alongside internal reports. The walkthrough demonstrates querying regional sales performance from structured data while referencing strategic priorities from unstructured PDFs in the same conversation. Additionally, the agent can generate formatted artifacts like PowerPoint presentations and Excel spreadsheets using natural language prompts. Security is maintained through least-privilege access controls, document-level ACLs, and built-in AI guardrails for content safety. Finally, the guide covers sharing agents with team members based on permissions and provides instructions for cleaning up resources to avoid ongoing charges.
aws