Open-Source Receipt Sorter Runs Locally, Keeping Client Invoices Off Third-Party Servers

A developer has built a receipt sorter that uses a local open-weight Gemma model served through Ollama to read photos, PDFs, and text files. It extracts the vendor, date, total, currency, and category, then copies each file into a sorted folder and produces a summary spreadsheet for accountants. The tool targets freelancers and small-business owners who want to avoid uploading sensitive client invoices to cloud receipt services that charge subscriptions. Because inference runs entirely on the user's laptop, no data leaves the machine and the tool works offline. The project is a single Python file that needs only Python 3.9 or later, with pypdf optional for PDF support.