Every week an Excel file lands in my inbox straight from QuickBooks. Each row holds a job number and a set of raw, “rough-draft” technician notes. Right now I copy those notes into ChatGPT to have them expanded and clarified for the customer, paste the polished text back into the sheet, jump into QuickBooks, locate the matching job or project, drop the text into its description field and finally create the invoice. The cycle is accurate yet painfully manual. I need that entire chain to run on its own. Here is what must happen: • As soon as the weekly spreadsheet arrives (or is dropped into a watched folder), it is parsed and each line of raw notes is sent to ChatGPT via the API. • ChatGPT returns an expanded, customer-friendly explanation of the work performed; grammar, spelling and clarity must be perfect. • The refined note is pushed back into the correct job or project in QuickBooks, matched by the job number already in the sheet. • Once every note is in place the corresponding invoice is generated and marked ready to email. No human touchpoints – I chose “Fully automated – no manual intervention.” The incoming notes have zero pre-formatting, so your workflow must be able to cope with free-form text and still give ChatGPT the right prompt each time. I am platform-agnostic: script it in Python, build a small serverless function, or wire it through Zapier/Make as long as it is robust, logs errors, and can be handed over with clear setup instructions. Deliverables I expect: • Working source code or automation blueprint. • One-time setup assistance and a short hand-off video or document. • A simple logging or alert system so I know if a note fails to post or an invoice fails to build. If you have previously connected Excel, QuickBooks API and OpenAI, that will move you to the top of the list. Let me know how you would structure the prompt that tells ChatGPT to expand the technician notes, and roughly how long the build will take.