What Happened
A developer has released TaxHacker, a self-hosted, open-source AI application designed to automate accounting tasks by processing receipts and invoices. The tool is built to run locally via a single Docker command, keeping all user data private.
What It Does
According to the announcement, TaxHacker performs the following core functions:
- Document Processing: Users can upload photos or PDFs of receipts and invoices. The app supports documents in any language.
- AI-Powered Data Extraction: An AI model (specific model not named in the source) extracts key fields including product names, amounts, taxes, dates, and merchant information.
- Currency Conversion: Automatically converts extracted amounts using historical exchange rates, with support for cryptocurrencies like Bitcoin (BTC) and Ethereum (ETH).
- Search & Management: Provides full-text search across all processed documents.
- Prompt Customization: Allows users to customize every Large Language Model (LLM) prompt, including system prompts, offering control over the extraction logic.
- Export: Data can be exported to a CSV file with all original documents attached for accountant review.
Technical & Licensing Details
The project is published on GitHub under the MIT License, making it 100% open-source. The primary deployment method is via Docker, emphasizing a one-command setup for ease of use. The core value proposition is data privacy; by being self-hosted, all processing occurs on the user's own infrastructure, and no data is sent to external servers.
Context
TaxHacker enters a space occupied by both commercial SaaS products (like Dext, Expensify, and Ramp) and other open-source projects. Its differentiation lies in the combination of local deployment, support for cryptocurrency transactions, and the unusual feature of fully customizable LLM prompts, which is typically a backend detail hidden from users in commercial offerings.
