This template creates an AI-powered movie recommendation chatbot that combines OpenAI’s GPT models with Qdrant vector search. It automatically loads IMDB’s Top 1000 movies from a GitHub repository, embeds their descriptions, stores them in a vector database (Qdrant), and responds to user queries with personalized movie suggestions.
Users can chat with the bot, provide examples of what they like (positive) and dislike (negative), and receive top-3 recommendations based on semantic similarity — all orchestrated within n8n.
Features
- Automated Data Ingestion: Fetches a CSV of IMDB’s Top 1000 movies from GitHub and extracts metadata (name, release year, description).
- Vector Embedding & Storage: Uses OpenAI embeddings (text-embedding-3-small) to create vector representations and stores them in a Qdrant collection.
- Intelligent Chatbot Agent: An OpenAI GPT-based agent (gpt-4o-mini) processes chat inputs, manages conversation context with memory, and uses a custom tool to fetch movie recommendations.
- Personalized Recommendations: Accepts both positive and negative examples from users and queries Qdrant’s recommendation API to return the top 3 most relevant movies.
- Flexible Integration: The chatbot can be triggered manually for testing or via chat messages, and the recommendation workflow can be reused as a standalone n8n tool.
- Clear API Interactions: Includes nodes for querying Qdrant for recommendations, retrieving movie metadata, and formatting results for user-friendly output.
