🥗 Nutrition Tutor Bot

AI-Powered Conversational Nutrition Assistant with Advanced RAG and Multimodal Capabilities

🧠 RAG System 📷 Multimodal AI 💬 Conversational ⚡ 10ms Search
Try Live Demo View on GitHub

Key Features

💬

Conversational AI Coach

Natural ChatGPT-style interface specialized for nutrition with context memory and personalized responses backed by evidence-based science.

🧠

Advanced RAG System

66 nutrition documents with 10ms vector search, semantic retrieval, and intelligent context selection for accurate, sourced responses.

📷

Multimodal Integration

Upload meal photos for AI-powered nutrition analysis using GPT-4 Vision with USDA API integration for real-time nutritional data.

🔍

Smart Database Search

AI-powered similarity search through comprehensive nutrition knowledge with instant results and detailed nutritional breakdowns.

🎯

Personalized Meal Planning

Goal-based meal recommendations with dietary restrictions, macro tracking, and customized nutrition strategies.

📊

Performance Analytics

Real-time metrics, conversation insights, and performance monitoring with 100% reliability and professional error handling.

System Architecture

User Interface Layer

💬 Chat Interface

  • Natural conversation flow
  • Context-aware responses
  • Real-time interaction

📊 Analytics Dashboard

  • Performance metrics
  • Usage statistics
  • System health monitoring

🔍 Database Search

  • Semantic similarity search
  • Instant results
  • Filtered nutrition data

📷 Photo Upload

  • Image processing
  • Food identification
  • Nutritional analysis

Core AI Engine

🧠 Conversational AI Coach

  • Context management
  • Personalized responses
  • Conversation memory
  • Dynamic prompting

📷 Multimodal Processor

  • GPT-4 Vision integration
  • Food identification
  • Portion estimation
  • Cross-modal analysis

🔍 RAG Query Engine

  • Query processing (10.1ms)
  • Vector similarity search
  • Context retrieval
  • Response generation

Data & Storage Layer

📚 Vector Database (ChromaDB)

  • 66 nutrition documents
  • Vector embeddings
  • Metadata storage
  • Semantic search index

🤖 OpenAI GPT-4

  • Text generation
  • Vision processing
  • Conversational AI
  • Context understanding

🍎 USDA FoodData Central

  • 400K+ food database
  • Real-time nutrition data
  • Comprehensive nutrients
  • Verified food information

📊 Local Processing

  • Sentence transformers
  • Local embeddings
  • Privacy protection
  • Fast vector operations

Data Flow Architecture

User Input
→
Query Analysis
→
Vector Search
→
AI Generation
→
Response

Performance Metrics

10.1ms Vector Search Time
66 Nutrition Documents
100% Reliability Score
12.5s RAG Pipeline Time
400K+ USDA Food Database
3 Core AI Components

Interactive Demo

Live Demo Coming Soon!

The interactive Streamlit demo will be deployed here. In the meantime, you can:

Clone & Run Locally

Usage Examples

💪 Muscle Building

User: "I want to build muscle"
Coach: "For muscle building, you'll need 1.6-2.2g protein per kg body weight daily..."

🥗 Meal Planning

User: "Create a vegetarian weight loss meal plan"
Coach: "Here's a balanced approach focusing on plant proteins..."

📷 Photo Analysis

User: [Uploads meal photo]
Coach: "I can see grilled chicken, rice, and vegetables. Here's the nutritional breakdown..."

Development Team

AB

Ashwin Badamikar

Student ID: 002055743

Role: RAG System Development, Conversational AI, API Integration, Testing & Performance, Frontend Development

Contact: badamikar.a@northeastern.edu

MA

Madhura Adadande

Student ID: 002306240

Role: Knowledge Base Design, Data Processing, Prompt Engineering, Multimodal Integration, Quality Assurance

Contact: adadande.m@northeastern.edu

Project Documentation

📚

GitHub Repository

Complete source code, setup instructions, testing scripts, and comprehensive documentation.

View Repository
🏗️

Technical Architecture

Detailed system architecture, RAG pipeline implementation, and multimodal integration design.

View Architecture
⚙️

Setup Instructions

Step-by-step installation guide, API configuration, and troubleshooting information.

Setup Guide
📄

MIT License

Open source project under MIT License for academic and commercial use.

View License

Academic Context

This project was developed for the Prompt Engineering and GenAI course at Northeastern University. It demonstrates mastery of advanced generative AI technologies including RAG systems, multimodal integration, and conversational AI.

Course: Prompt Engineering & GenAI Institution: Northeastern University Year: 2025