Expense tracking system that converts voice or text input into structured accounting records. Users can submit expenses through a web interface or Discord bot. The system transcribes voice messages, categorizes expenses using machine learning and vector similarity, and stores them for reporting and export.
Key features
Web dashboard for manual expense entry and voice recording
Discord bot as alternative submission interface
Whisper transcription → GPT-3.5 parsing → ML classification pipeline
Qdrant vector search for semantic similarity between expenses and categories
User corrections stored as training data for periodic classifier retraining