Automated File Monitoring and AI Processing Workflow

This n8n workflow automates the monitoring and processing of files within a designated local folder, specifically tailored for handling bank statements or similar documents. Its primary goal is to synchronize local files with a vector database (Qdrant), enabling advanced retrieval and AI-powered Q&A interactions on the documents. The process begins with a local file trigger that detects added, changed, or deleted files in a specified directory. When a file is added, it is read and processed to create embeddings via Mistral Cloud, then stored in Qdrant for future retrieval. If a file is deleted, its corresponding vector point in Qdrant is also removed to keep the database updated. For file updates, the existing vector point is deleted and re-created with the latest data to ensure accuracy. Additionally, the workflow integrates a LangChain-based assistant that can respond to user queries related to the stored documents, leveraging both the vector store and advanced AI chat models. This setup is ideal for organizations managing large sets of documents who want automated, real-time indexing, and AI-enabled query capabilities for efficient information retrieval.