A practical starter kit for finance and operations leaders — plus what to ask once the discussion gets going. TL;DR In Databox’s Time to Insight study, 64% of teams report that it takes them one to three days to pull together the data needed to answer a business question, and 73% say their biggest reporting headache is data being spread across multiple sources. Finance teams feel this more than most: a single question can span Stripe, QuickBooks, and a spreadsheet. This article shares nine real prompts from the Databox Prompt Library (five for Stripe, four for QuickBooks), each with an example response, why it matters, and two to three natural follow-up questions. Prompts are effective when they specify the metric, the time frame, and the comparison. “How’s my revenue?” doesn’t cut it. “What is our MRR this month, and how has it trended over the last 6 months?” does. Cross-source questions (like margin vs. MRR growth, or revenue vs. cash-position gaps) only become answerable once Stripe and QuickBooks sit in a unified layer. Databox manages the data connection; Genie or MCP manages the querying. The key architectural distinction from generic AI tools: with Genie, the language model never performs the calculations. All math runs in deterministic code against validated data. That’s the difference between trustworthy BI and AI that merely approximates arithmetic. Introduction You’re twenty minutes into chasing down a number that should take thirty seconds. Net new MRR. Last quarter’s gross profit margin. Cash on hand before Friday’s payroll. The data is there. Stripe has part of it. QuickBooks has the rest. The AI tools open in your other browser tab are…