AI-powered personalized nutrition assistant built with Next.js 15 and the Vercel AI SDK.
Personalized sidebar with diet type, goals, calorie targets, and quick action prompts.
Conversational meal planning with follow-up questions to tailor recommendations.
Detailed 7-day meal plans with per-meal calorie and macro breakdowns.
Complete weekly plans with batch cooking strategy and grocery lists.
NutriBot is an intelligent nutrition chatbot that provides:
- Personalized meal recommendations based on your diet (vegan, keto, paleo, Mediterranean, etc.)
- Real nutrition data — accurate calories and macros fetched from the Nutritionix API
- Agentic AI — the LLM autonomously calls nutrition tools to look up meal data before responding
- Streaming responses — answers appear in real-time as they are generated
- Persistent preferences — diet type, goals, and calorie targets saved to localStorage
| Layer | Technology |
|---|---|
| Framework | Next.js 15 (App Router) |
| Language | TypeScript |
| UI | Tailwind CSS v3 + Framer Motion |
| State | Zustand with localStorage persistence |
| AI SDK | Vercel AI SDK v6 |
| Nutrition Data | Nutritionix API |
| Deployment | Vercel |
src/
├── app/
│ ├── api/
│ │ ├── chat/route.ts # Streaming endpoint (agentic loop)
│ │ └── nutrition/route.ts # Nutritionix proxy (API key secured server-side)
│ ├── chat/page.tsx # Main chat UI
│ ├── layout.tsx
│ └── globals.css
├── components/
│ ├── ChatInterface.tsx # Root layout: sidebar + chat
│ ├── MessageBubble.tsx # Streaming message with markdown
│ ├── MessageInput.tsx # Auto-resizing input with stop button
│ ├── NutritionCard.tsx # Macro visualization cards
│ ├── QuickActions.tsx # Conversation starters
│ ├── Sidebar.tsx # Diet prefs, goals, calorie target
│ └── TypingIndicator.tsx
├── hooks/useNutritionChat.ts # Wraps Vercel AI SDK useChat
├── lib/
│ ├── nutritionix.ts # Vercel AI SDK tool definition
│ ├── system-prompt.ts # Dynamic system prompt
│ ├── types.ts
│ └── utils.ts
└── store/chatStore.ts # Zustand store (persisted)
The API route at /api/chat uses streamText from the Vercel AI SDK with maxSteps: 3, enabling an agentic loop where the AI can:
- Decide to search for nutrition data
- Call the
searchNutritiontool (backed by Nutritionix API) - Receive the results and craft a response with accurate macros
- Stream the final answer to the user
- Node.js 18+
- An Anthropic API key
- (Optional) Nutritionix API credentials
# 1. Clone the repo
git clone https://github.com/vinaybudideti/Masters-Project.git
cd Masters-Project
# 2. Install dependencies
npm install
# 3. Configure environment
cp .env.example .env.local
# Edit .env.local and add your ANTHROPIC_API_KEY
# 4. Start dev server
npm run devOpen http://localhost:3000.
Note: Without
NUTRITIONIX_APP_IDandNUTRITIONIX_API_KEY, the app uses estimated mock nutrition data. The chat still works fully — the AI uses its built-in knowledge for nutrition estimates.
- Push to GitHub
- Go to vercel.com → Import Project → select your repo
- Add environment variables:
ANTHROPIC_API_KEY(required)NUTRITIONIX_APP_ID(optional)NUTRITIONIX_API_KEY(optional)
- Click Deploy
Your app will be live in ~60 seconds at a *.vercel.app URL.
- Streaming AI responses — text streams in real-time via SSE
- Agentic nutrition lookup — the AI calls the Nutritionix API autonomously when it needs accurate data
- Inline nutrition cards — animated macro cards appear in chat alongside AI responses
- Persistent preferences — diet type, health goals, calorie targets, and restrictions saved across sessions
- Dark glassmorphic UI — modern design with smooth Framer Motion animations
- Mobile responsive — collapsible sidebar for mobile
- Zero-config fallback — works without Nutritionix keys (uses estimated data)
The original Rasa NLU + Flask implementation is preserved in the legacy/ directory for reference:
legacy/rasa/— Rasa 3.6 NLU chatbot (training data, models, actions)legacy/flask/— Flask REST API middleware
The new architecture replaces both with Next.js API Route Handlers and the Vercel AI SDK.
MIT © 2025 vinaybudideti