Automated OCR Receipt Processing to Notion

This n8n workflow automates the extraction and processing of receipt data using OCR, and seamlessly integrates the information into Notion. It starts with a webhook trigger to receive receipt images, then uses a language model to split and embed the text. The embeddings are stored in a vector database (Supabase), which is queried for related data. A retrieval-augmented generation (RAG) agent uses this information to understand and handle the receipt data, ultimately logging the status in Google Sheets. If an error occurs, a Slack alert notifies the user of the issue.

The workflow involves key nodes such as the webhook trigger, text splitter, OpenAI embeddings, Supabase vector store, a language model for contextual understanding, and Google Sheets for logging. It is ideal for automating receipt management, expense tracking, or any scenario where receipts are digitized and logged automatically, saving time and reducing manual effort.