I photograph music scores for my eBay.co.uk store and want those images to become live listings without the usual copy-and-paste routine. The job is to build a tool that scans a photo, pulls out the visible text with reliable OCR or similar vision tech, then uses that information to fill the Title and Description and SKU fields of a new eBay listing. These photos will all be attached to the listing too, apart from the final, SKU photo (see later for details). I have attached two sample sets of photos, the first photo currently shows the outside of the book, this may or may not have all the necessary text visible, the second photo will be the publisher's page which should have all the necessary information in. I can change the order we take the photos in if necessary. The information I want extracted needs to be laid out as follows: COMPOSER: TITLE: PUBLISHER: ANY OTHER DETAILS. There needs to be an input box for a title suffix, depending on what we're listing we need to add a certain ending on to the title, truncating the other text as ncessary, this might be :MUSIC SCORE or :VOCAL SCORE but we need to be able to change this and enter it before batch listing. I've attached a photo of an imagined interface. See linked current example on our store at the moment - https://www.ebay.co.uk/itm/297601453750 The final photo will always be the same format and is our locator slip which we use when picking sold items, it contains the Custom Label (SKU) which is on the bottom two lines, the first part (second to bottom line of text) takes the form of the shelf location (such as V02-B03 or X01-H05) and then the very bottom line is the item number (such as 002 or 104) this needs to be entered into the custom label field as one string as follows: V02-B03-002. This will always be the final photo for that listing so the next photo is where to begin the next listing. Price: I would like to be able to enter the price for a batch. Description I need to have a text input for each batch. Any other eBay fields like postage policy etc. will be standard and can be populated from a template or standard directions on set-up. If this is best in the interface from drop-down boxes or data input then that is fine too. The process has to be hands-off: I drop the photos into a folder (or an S3 bucket, Google Drive, etc.), open the program, enter the details I need and the script calls the eBay Sell/Trading API to create a live listing on the UK site, already populated with the extracted text, I can cope with it being a browser run interface as long as there are no issues with eBay flagging it as bot-run. Python with Tesseract or OpenCV is fine, as is an equivalent approach in Node, but whatever you choose must run on Windows 11 and be easy for me to maintain. Clarity of code and clear mapping from “text found” → “field filled” is essential. Deliverables • Source code with concise README • Setup guide for local or cloud deployment • Short demo (screen capture or live session) showing at least one photo automatically turned into a populated eBay draft listing on eBay.co.uk Once I can feed a photo through your tool on my local PC and see everything appear correctly on a listing, the job is complete.