Extract Expense Data from a Receipt into CSV
Use this prompt after uploading a single receipt image or PDF to ChatGPT or Claude. It extracts structured expense data from one document and returns it as a CSV row ready for pasting into a spreadsheet.
The Prompt
You are processing a business expense receipt. Extract the following fields and return them as a single CSV row with these headers: - Date (YYYY-MM-DD) - Vendor Name - Total Amount (numbers only, no currency symbol) - Currency - Payment Method Last 4 Digits (if visible, otherwise leave blank) - Expense Category (use IRS categories: Travel, Meals, Office Supplies, Software Subscriptions, Advertising, Professional Services, Utilities, Vehicle, Other) - Business Purpose (meals only: note the client or meeting purpose if visible; otherwise leave blank) - IRS Compliance Flag (flag YES if: the receipt is for meals without a business purpose noted, the expense is over $75 and no receipt documentation is present, or the vendor name is unclear) Return only the CSV row. No explanation. No headers — I will add those separately.
From the guide
How to use AI to turn a messy inbox of supplier invoices and receipts into a clean expense report without a bookkeeper →Related Prompts
Simulate a Tough Banker Interview With AI Feedback
Use this prompt after reviewing your Q&A prep sheet to simulate a real banker interview. Answer each question as you would in the actual meeting, and the AI will give you specific feedback on what landed well and what sounded weak or vague.
Rewrite a Loan Brief in Plain Business Language
Use this prompt after reviewing the AI-drafted loan brief if the language feels overly formal or generic. It produces a version that sounds more like you, which is important since you'll be handing it to a loan officer with your name on it.
Model Alternative Loan Structures for a Weak DSCR
Use this follow-up prompt if your initial DSCR calculation comes back between 1.0 and 1.25. It helps you model alternative loan structures and prepares you to address a weak coverage ratio before the lender raises it.
Prepare a Full Bank Loan Application Brief
Use this prompt at the start of your bank loan meeting prep session. Paste in your raw financial data — 12 months of revenue, monthly expenses, existing debt payments, loan amount requested, intended use of funds, and business assets — to get a DSCR calculation, a 5 Cs evaluation, a one-page loan brief, and a Q&A prep sheet in one response.