One chat agent for both your sales spreadsheet and your strategy PDFs: put structured data and document knowledge in the same workspace so answers cite both and nobody has to know where the number lived: AI agent prompt
An AWS Machine Learning Blog how-to (Nov 3, 2026) shows Amazon Quick combining an enriched sales dataset (text-to-SQL) with an S3 knowledge base of business PDFs inside one Space, then linking a custom chat agent that answers across both with citations. Example questions span “How did sales change by region?” and “How do 2025 results align with our market analysis memo?” The walkthrough covers semantic column descriptions, custom dataset instructions, Spaces, persona instructions, sharing with least-privilege access, and generating PowerPoint/Excel/Word from chat. Needs an Amazon Quick Enterprise subscription for dataset creation; adapt the same pattern on any stack that can join warehouse metrics with a document store behind one agent.
The prompt
1. Pick one spreadsheet or warehouse extract (sales, pipeline, support volume) and three trusted PDFs/memos that explain strategy or definitions. 2. Enrich the table: correct types, write a dataset description, usage rules (currency, fiscal year, how Sales is calculated), and a one-line description per important column. 3. Index the documents in a knowledge base and put the dataset + knowledge base in one shared workspace your chat agent can see. 4. Ask three questions that need both sources (trend from the numbers + “why” from a memo) and require citations on every figure. 5. Share only after access matches who may see those docs; keep a human check before any slide or number leaves the company.
How to use it
- Copy the prompt above.
- Replace every [bracket] with your own details (or fill in the blanks on AgentPlaybook Real Estate, which remembers your answers for the next prompt).
- Paste it into Amazon Quick (Spaces, datasets, knowledge bases and review the answer before you use it.
Works with: Amazon Quick (Spaces, datasets, knowledge bases, chat agents) or similar BI+RAG agent stack, S3 or document store, spreadsheet/warehouse extract
Real-world example: aws.amazon.com/blogs/machine-learning/bring-structured-and-unstructure
Featured in: AI prompts for Google Sheets and Excel work · AI prompts for writing SOPs, reports, policies and checklists
Before you use it: what to check
Already built into this prompt
- Share only after access matches who may see those docs; keep a human check before any slide or number leaves the company.
- Ask three questions that need both sources (trend from the numbers + “why” from a memo) and require citations on every figure.
Check yourself
- Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details.
Using an AI agent? Download this prompt as a SKILL.md agent skill with these checks built in, or get the .zip for Claude.ai. How to install skills.
Want to make this prompt even better? Grade it with the free prompt grader or build your own with the AI prompt generator.
32 copy-paste agent prompts (PDF) plus 5 new prompts every week.
Get it free