Transform Gmail Emails into Vector Embeddings with PGVector and Ollama

This workflow efficiently processes your Gmail emails by storing them in both structured and vectorized formats. It uses Ollama as the LLM provider and PGVector as the vector database engine, running locally on PostgreSQL. The workflow initially requires a manual trigger to import historical emails in monthly batches. Once activated, it continuously updates with new emails via the Gmail Trigger. This dual storage approach allows for both semantic searches using vector similarity and precise data retrieval through structured queries.