HelloFresh Weekly Recipe Recommendation - n8n Automation Template

This workflow builds a personalized recipe recommendation engine using the weekly HelloFresh menu. It scrapes current recipes, extracts structured data, enriches it with embeddings, and stores it in both a Qdrant vector database and SQLite. An AI Agent, powered by Mistral Cloud and Qdrant’s recommendation API, allows users to chat and get tailored recipe suggestions based on their likes, dislikes, dietary preferences, and cooking needs.

Features:

• 🔄 Weekly Menu Fetching – Automatically scrapes HelloFresh’s current week menu.
• 📝 Recipe Data Extraction – Collects recipe name, description, ingredients, utensils, instructions, nutrition, and tags.
• 📦 Data Storage – Saves recipes in both SQLite (structured storage) and Qdrant (vector search).
• 🧠 Embeddings & Vectorization – Uses Mistral embeddings for semantic search.
• 🤖 AI Chat Agent – Lets users interact with an AI to request personalized recommendations.
• 🎯 Advanced Recommendations – Leverages Qdrant’s Recommend API with positive & negative queries to fine-tune results.
• 🔍 Unique Results – Groups recipes to avoid duplicates and ensure diverse recommendations.
• 💬 Interactive Chatbot – Suggests recipes based on taste (e.g., spicy, comforting, quick meals) and restrictions (e.g., allergens, disliked flavors).