Email Data Processing with Vector Embeddings in n8n

This n8n workflow automates the process of extracting, enriching, and storing email data for advanced analysis. It begins with a manual trigger or scheduled Gmail polling, retrieves emails within specified date ranges, and extracts key email fields such as sender, recipient, subject, and text. The emails are then stored in a PostgreSQL database as structured metadata. Simultaneously, the email content is split into manageable chunks, embedded into vector representations using Ollama’s language model, and stored in a vector database for similarity searches. The process includes dynamic date calculations to fetch emails over specified weekly intervals, making it suitable for ongoing email analysis, customer support, or research projects involving large email datasets.