Intelligent Product Recommendation System Using n8n

This workflow creates an advanced product recommendation system leveraging n8n’s automation capabilities combined with AI and vector databases. It is designed to offer personalized product suggestions based on user interactions and product data. The process begins with a manual trigger, allowing users to initiate the workflow, such as when they need updated recommendations.

The workflow involves retrieving product data from a Google Sheets database, generating embeddings using OpenAI’s language models, and storing these vectors in a Qdrant vector database for efficient similarity search. When a user sends a chat message, the workflow triggers a process that fetches relevant product embeddings, processes the message with OpenAI’s large language model, and employs a Retrieval-Augmented Generation (RAG) approach to generate personalized suggestions.

The system splits product descriptions into manageable chunks for better embedding accuracy, loops over each product to process and store data, and then uses the stored vectors for real-time similarity searches when a chat is received. This setup enables highly relevant, dynamic product recommendations based on user queries, making it ideal for e-commerce sites or customer support platforms seeking personalized experiences.