An intelligent WhatsApp chatbot powered by an AI backend to provide relevant and contextual responses. This project uses a two-part architecture: a WhatsApp bot frontend built with Node.js and an AI backend built with Python and FastAPI.

Repositori ini juga berisi tugas akhir dengan judul:

Pengembangan Chatbot AI Berbasis Large Language Model (LLM) LangChain dengan Retrieval-Augmented Generation (RAG) pada Customer Service di PT. Telemedia Prima Nusantara

This project consists of two main services working together:

1. WhatsApp Bot (Node.js):

   • Located in the whatsapp-bot/ directory.
   • Uses the Baileys library to connect to WhatsApp.
   • It's responsible for receiving incoming messages, sending replies, and handling basic interactions like greetings or static info.
   • For complex questions, the bot forwards them to the AI backend via an API request.

2. AI Backend (Python):

   • Located in the root directory.
   • Built with FastAPI to provide an API endpoint (/chat).
   • Uses LangChain and models from OpenAI to understand and process natural language questions.
   • Leverages FAISS CPU as a vector store for relevant information retrieval (Retrieval-Augmented Generation - RAG) from available data.

• Quick Greeting Responses: Handles common greetings directly at the bot level for efficiency.
• Static Info: Provides quick answers for common questions like addresses or technician contacts.
• AI Processing (RAG): Forwards complex questions to the Python backend for context-aware, data-driven answers.
• History Logging: Saves conversation history for each user.
• Media Handling: Saves media files sent by users.

Component	Technology
WhatsApp Bot	Baileys, Axios, Pino
AI Backend	LangChain, OpenAI, FAISS, Uvicorn
Deployment

Ensure you have Python 3.8+ installed.

1. Clone the repository:

   git clone https://github.com/akariwill/Chatbot-AI.git
   cd chatbot-ai

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install Python dependencies:

   pip install -r requirements.txt

4. Set up environment variables:

   • Create a .env file in the root directory.
   • Add your OPENAI_API_KEY to this file:

   OPENAI_API_KEY="sk-..."
   

5. Run the FastAPI server:

   uvicorn app:app --host 0.0.0.0 --port 8000

   The AI server is now running at http://localhost:8000.

Ensure you have Node.js v16+ installed.

1. Navigate to the bot directory:

   cd whatsapp-bot

2. Install Node.js dependencies:

   npm install

3. Run the bot:

   npm start

   • A QR code will appear in the terminal.
   • Scan the QR code with your WhatsApp mobile app (Link a device).

Once both services are running, your bot will be active and ready to respond to messages on WhatsApp.

chatbot-ai/
├── index.html              # Website tugas akhir (isi skripsi dalam HTML)
├── Research.pdf            # Dokumen lengkap skripsi (PDF)
├── extract_images_from_pdf.py website
├── website/               
│   ├── css/
│   ├── javascript/
│   └── img/
│       └── research/       
├── whatsapp-bot/           # WhatsApp Bot Frontend (Node.js)
│   ├── index.js
│   ├── package.json
│   └── auth_info/         # (Auto-generated, ignored by Git)
├── app.py                  # Entry point for the AI server (FastAPI)
├── main.py                 # Core chatbot logic (LangChain)
├── requirements.txt       # Python dependencies
├── Dockerfile              # Configuration for containerization
├── assets/
│   └── Data/               # Data for RAG
└── ...

This project is licensed under the MIT License.

If you have any questions or feedback, feel free to contact me on Discord wildanjr_ or Instagram akariwill.