Governed reports with Amazon Quick Desktop and Amazon FSx for NetApp ONTAP
Amazon Web Services Ebbey Thomas
Amazon Quick Desktop can turn FSx for ONTAP files into weekly reports in minutes, not hours. The catch: it keeps the source docs governed and cited, so the AI doesn’t get to freestyle.
Based on reporting by Amazon Web Services, Ebbey Thomas — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Amazon Web Services is pushing Amazon Quick Desktop into a very specific job: turning a governed pile of business files into weekly reports without losing track of where the facts came from. The pitch is simple enough. What used to take hours each week — rereading documents, reformatting metrics, copying summaries into Slack — becomes a guided workflow that runs from the desktop.
The setup ties Amazon Quick Desktop to Amazon FSx for NetApp ONTAP, an Amazon S3 access point, and Slack. A team chooses one approved reporting folder, keeps drafts and restricted files out of it, and lets Amazon Quick build a searchable knowledge base from that narrow slice of content. The example knowledge base is called Business Reporting Archive. That’s the source of truth the assistant uses for citations, reports, deck drafts, visuals, and Slack-ready summaries.
There’s a lot of discipline baked into the design, and that’s the point. Amazon Quick only gets read access to the approved prefix through IAM and the access point. Human review stays in place before anything goes to Slack. Even the reporting assistant itself is framed as a reusable skill called Weekly Business Reporting Assistant, with a trigger phrase and instructions for how to answer, cite, and ask for clarification.
Amazon also uses the desktop app’s personal knowledge graph to connect people, projects, events, actions, and documents. But the Business Reporting Archive remains the governed source for report facts. The pattern is aimed at weekly business reviews, operating plans, forecast updates, risk reporting, and executive briefings — basically any place where leadership wants speed, but not at the cost of traceability.
My take — AI-written commentary, not fact-checked reporting
This is the right kind of AI product: narrow, governed, and boring in all the useful ways. The industry keeps selling magic; the real win is making reporting less miserable without turning Slack into a rumor mill with citations. Open-ended chat is fine for brainstorming. For business reporting, controlled inputs and human approval are the whole game.
Read more about this at: Amazon Web Services