Recipe Recommendations with Qdrant & Mistral AI

This n8n workflow builds a recipe recommendation system by scraping HelloFresh's weekly menus, processing recipes into chunks, generating embeddings with Mistral AI, and storing them in a Qdrant vector database alongside full details in SQLite. It enables an AI Agent to deliver tailored recipe suggestions via Qdrant's Recommendation API, which supports negative prompts to exclude unwanted ingredients or recipes, enhancing user satisfaction.

Key benefits include scalable recommendation features for AI agents, easy customization for other domains like clothing or vehicles, and integration of advanced vector search for precise, context-aware results. It saves significant development time by automating data ingestion, embedding, storage, and querying.

Ideal use cases: e-commerce product recommendations, content personalization in apps, or customer service bots. Requires Qdrant instance and Mistral.ai account; perfect for teams building intelligent recommendation engines.