AI home recognition using photo with address and metadata

Project Overview I'm a real estate investor. My team takes exterior photos of homes in the field. I need a tool that turns a Google Drive folder of those photos into a clean address list I can export as CSV. How it should work I upload/import batches of home exterior photos into a Google Drive folder The tool reads each photo's GPS metadata (EXIF) and reverse-geocodes it to get street name, city, and ZIP — GPS alone isn't reliable for the exact house number The tool then uses AI vision / OCR on the photo itself to read the house number from the home (curb, door, mailbox, etc.) If no house number is visible or readable, that field is marked "Not Available" — never guessed When a batch is processed, I can export a CSV with: photo filename, house number, street, city, ZIP, and a confidence/status column Requirements Works with Google Drive as the input source (batch processing, not one-by-one) Handles 100+ photos per batch without manual babysitting Accurate reverse geocoding (Google Maps API or similar) Vision extraction for house numbers (GPT-4 Vision, Claude, Google Vision, or similar) Simple to run — I'm not a developer; a button, script, or simple web page is fine CSV export I can open in Excel/Sheets