Skip to content

Latest commit

 

History

History
163 lines (118 loc) · 5.8 KB

README.md

File metadata and controls

163 lines (118 loc) · 5.8 KB

Moonflour_Bean – AI Chatbot for Coffee Shop Apps

Welcome to the Moonflour_Bean Chat Agent repository! This project contains the full-stack implementation of an AI-powered Multi-agent app designed to enhance customer experiences in a coffee shop application by simulating real-world interactions.

The chatbot utilizes Large Language Models (LLMs), Natural Language Processing (NLP), RunPod’s infrastructure, and Firebase to assist customers with:

✅ Taking orders
✅ Answering detailed menu-related queries
✅ Providing personalized product recommendations

This chatbot is integrated within a React Native mobile app (currently under development).

🎯 Project Goals

The goal of this project is to develop an intelligent, agent-based chatbot capable of:

  1. Handling real-time customer interactions – Allowing customers to place orders seamlessly.
  2. Answering menu-related questions – Providing details on ingredients, allergens, and other menu information using a Retrieval-Augmented Generation (RAG) system.
  3. Offering personalized recommendations – Utilizing a market basket analysis recommendation engine to suggest complementary items.
  4. Ensuring structured order placement – Guiding customers through an efficient ordering process for accuracy.
  5. Filtering inappropriate queries – Using a Guard Agent to block irrelevant or harmful inputs.

🛠️ Core Agents in the Chatbot System

1. Guard Agent (Content Filtering)

  • Serves as the first line of defense by monitoring all incoming queries.
  • Blocks inappropriate, harmful, or irrelevant messages before they reach other agents.
  • Ensures safe and meaningful interactions with the chatbot.

2. Order Taking Agent (Order Assistance)

  • Guides users through the entire order placement process using chain-of-thought reasoning.
  • Captures all customer preferences and order details in a structured, logical manner.
  • Ensures orders are complete and accurate before processing.

3. Details Agent (RAG System) (Menu Information)

  • Uses Retrieval-Augmented Generation (RAG) to provide detailed answers on menu items, including:
    • Ingredients
    • Allergens
    • Nutritional information
  • Retrieves relevant data from a vector database and generates accurate responses.

4. Recommendation Agent (Personalized Suggestions)

  • Works alongside the Order Taking Agent to suggest complementary products based on the user’s order history.
  • Uses a market basket analysis model to enhance upselling opportunities and help users discover new products.

5. Classification Agent (Query Routing)

  • Analyzes user intent and directs queries to the appropriate agent.
  • Ensures queries are handled efficiently, whether they relate to:
    • Order placement
    • Menu details
    • Product recommendations

⚙️ How the Agents Work Together

  1. User Query Processing:

    • A customer sends a message to the chatbot.
    • The Guard Agent filters out any harmful or irrelevant content.

  2. Intent Classification:

    • The Classification Agent determines the purpose of the query (e.g., ordering, menu details, recommendations).

  3. Task Delegation:

    • The query is sent to the most relevant agent:
      • Order Taking Agent for order-related queries.
      • Details Agent for menu-related questions.
      • Recommendation Agent for product suggestions.

  4. Upselling Process:

    • Before finalizing an order, the Order Taking Agent may request recommendations from the Recommendation Agent to enhance the customer’s experience.

📂 Folder Structure

coffee_shop_customer_service_chatbot/
│   ├── Frontend/                # React Native app code (Coming soon)  
│   ├── python/ (Backend)  
│       ├── API/                # Backend API for the agent-based chatbot system  
│       ├── dataset/            # Dataset for training the recommendation engine  
│       ├── products/           # Product details (names, prices, descriptions, images)  
│       ├── vector_database.ipynb  # Builds the vector database for the RAG model  
│       ├── firebase_uploader.ipynb  # Uploads product data to Firebase  
│       ├── recommendation_engine.ipynb  # Trains the recommendation system

🚀 Getting Started

Prerequisites

Ensure you have the following installed on your system:

  • Python 3.8+
  • Node.js & npm
  • React Native
  • Firebase CLI (for data management)
  • RunPod API Key (for cloud-based inference)

Installation

Backend Setup

# Clone the repository
git https://github.com/buriihenry/Moonflour_Bean.git

# Install dependencies
pip install -r requirements.txt

Frontend Setup

# Navigate to the frontend directory
cd ../Frontend

# Install dependencies
npm install

Running the Project

Start the Backend

cd python
python main.py

Start the Frontend

cd ../Frontend
npm start

📌 Contributing

Contributions are welcome! If you'd like to contribute, please follow these steps:

  1. Fork the repository.
  2. Create a feature branch (git checkout -b feature-name).
  3. Commit your changes (git commit -m 'Add feature').
  4. Push to your branch (git push origin feature-name).
  5. Open a pull request.

📜 License

This project is licensed under the MIT License – see the LICENSE file for details.

📞 Contact

For any inquiries or support, feel free to reach out: 📌 GitHub: buriihenry

Happy coding! ☕🤖