Health Tech · Nutrition & Wellness
AI-Powered Personalized Diet Plan Generator
Multi-agent AI system generating personalized diet plans in under 10 seconds using RAG over 1000+ indexed nutrition books.
- Role
- AI Developer
- Built multi-agent pipeline for conversational data collection and diet plan generation.
- Industry
- Nutrition & Wellness
- Timeline
- Completed
Business impact
- books indexed
- 1000+
- response time
- Under 10 sec
- retrieval
- Hybrid Vector
Overview
Built a multi-agent AI system that generates personalized diet plans through conversational data collection, intelligently gathering user preferences, health history, and dietary requirements.
The chatbot asks targeted questions until all required user data is gathered, then delivers complete end-to-end diet plans using RAG over 1000+ indexed nutrition books. Users can modify and enhance their generated diet plans based on preferences.
Technical architecture
- Multi-Agent System
- LangChain-based orchestration for intent handling and plan generation
- Vector Database
- Qdrant for hybrid retrieval across nutrition knowledge base
- RAG Pipeline
- Retrieval-augmented generation over 1000+ indexed documents
- Backend
- FastAPI endpoints for recommendations and user history management
Key features
- Conversational AI chatbot for targeted data collection
- Personalized diet plans based on user history and health goals
- Knowledge base of 1000+ nutrition and health books
- Real-time generation in under 10 seconds
- Plan customization and modification capabilities
- Evidence-based recommendations from indexed documents
Challenges solved
- Building conversational flow for comprehensive data collection
- Indexing and searching 1000+ nutrition books efficiently
- Achieving sub-10 second response time for complete diet plans
- Implementing hybrid retrieval with Qdrant vector database
Tech stack
- LangChain
- Multi-Agent Systems
- Qdrant
- FastAPI
- RAG
- Vector Databases
- Python
- LLM
- Conversational AI