🍛 MASala: Multi-Agent AI That Cooks Up Recipes Just for You
~ From fridge to feast, MASALA plans it all.
- 👨💻 Dhyey Joshi
- 👨💻 Ashutosh Gupta
- 👩💻 Vedika Shinde
- 👨💻 Naman Shah
Our Multi-Agent System intelligently assists users in discovering healthy, personalized recipes using real-time inputs such as available ingredients, dietary restrictions, and allergies. With agents collaborating as an Analyzer, Nutritionist, Chef, and Presenter, the system ensures safe, delicious, and nutrition-aware meals — all generated automatically using modern AI.
By leveraging AI agents to simplify healthy cooking at home, our system:
- Reduces cognitive load on users,
- Promotes food safety through allergy and dietary filtering,
- Encourages smarter, personalized eating habits.
It’s a modern take on accessible wellness — powered by collaborative intelligence.
MASala AI is a creative, multi-agent system that generates personalized recipes for users by coordinating a set of intelligent agents. It adapts to user preferences, dietary needs, and allergies, delivering healthy and delicious options.
Watch MASala AI in action!
👉 Click here to view the demo on YouTube
| Agent | Role |
|---|---|
| 🧪 Web Analyzer | Scrapes and analyzes trending recipes across the web. |
| 🥦 Nutritionist | Filters ingredients based on dietary restrictions and allergies. |
| 🍳 Chef | Crafts creative, personalized recipes based on filtered inputs. |
| 🗣 Presenter | Generates readable recipe JSON and visual prompts for sharing. |
All agents communicate through shared data, orchestrated via CrewAI, ensuring a seamless flow and conflict-free coordination.
- ⚙️ CrewAI – Orchestrates agent behavior and logic.
- 🧠 Gemini (Google AI) – Powers intelligent agent reasoning.
- 🔗 LangChain + LangSmith – Memory, observability, and traceable runs.
- 🎨 Pollinations API – Generates AI-based food imagery.
- 🚀 FastAPI – Powers the backend REST API.
- 💡 React.js – Interactive frontend dashboard.
git clone https://github.com/Naman009/MASala.git
cd MASalaCreate a .env file with the following:
LANGSMITH_TRACING=true
LANGSMITH_ENDPOINT="https://api.smith.langchain.com"
LANGSMITH_API_KEY="your_langsmith_key"
LANGSMITH_PROJECT="your_project_id"
GOOGLE_API_KEY="your_google_api_key"
SERPER_API_KEY="your_serper_key"python -m venv masala_env
source masala_env/bin/activate # Windows: masala_env\Scripts\activate
pip install -r backend/requirements.txtcd backend
uvicorn main:app --reloadcd frontend
npm install
npm run dev| Endpoint | Description |
|---|---|
POST /generate |
Send user input → Get back recipe JSON |
GET /logs |
View LangSmith logs of recent agent runs |
GET /trace |
Fetch public trace link for debugging |
- ✅ View active agents and their current tasks.
- 📡 Real-time communication and trace logs.
- 🍽️ Final JSON response and shareable recipe visual from Pollinations AI.
- 📊 Agent status: Working, Pending, Completed
This project is licensed under the MIT License. See the LICENSE file for details.