ChatSBT

ChatSBT is a full-stack chat application that allows users to interact with multiple AI models through a web interface. The application features real-time streaming responses, chat history persistence, and support for various language models.

Features

• Real-time chat with multiple AI models
• Server-Sent Events (SSE) for streaming responses
• Chat history persistence
• Responsive web interface built with Svelte
• Docker support for easy deployment

Technology Stack

Backend

• Python 3.11+
• Starlette - ASGI framework for building asynchronous web applications
• Langchain - Framework for developing applications with LLMs
• Masonite ORM - Database ORM for Python
• SQLite - Default database (configurable)

Frontend

• Svelte 5 - Reactive UI framework
• Vite - Next-generation frontend tooling
• DaisyUI - Tailwind CSS components
• Marked - Markdown parser

Infrastructure

• Deno - JavaScript/TypeScript runtime for frontend build
• uv - Python package installer and resolver

Prerequisites

Before you begin, ensure you have the following installed:

• Python 3.11 or higher
• Deno 2.4 or higher
• uv (Python package installer)
• Docker (optional, for containerized deployment)

Installation

Backend Setup

1. Install Python dependencies using uv:

   uv sync
   

2. Set up environment variables: Create a .env file in the project root with the necessary configuration:

   # Database configuration
   DB_CONNECTION=sqlite
   DB_DATABASE=chatsbt.db
   
   # AI Provider API keys (add as needed)
   OPENAI_API_KEY=your_openai_api_key
   ANTHROPIC_API_KEY=your_anthropic_api_key
   

Frontend Setup

1. Navigate to the frontend directory:

   cd frontend
   

2. Install frontend dependencies:

   deno install --allow-scripts
   

3. Build the frontend:

   deno run build
   

Running the Application

Development Mode

To run the application in development mode with hot reloading:

1. Start the backend server:

   uv run app.py
   

   The backend will be available at http://localhost:8000

2. In a separate terminal, start the frontend development server:

   cd frontend
   deno run dev
   

   The frontend will be available at http://localhost:5173

Production Mode

To run the application in production mode:

1. Build the frontend:

   cd frontend
   deno run build
   cd ..
   

2. Start the backend server:

   uv run app.py
   

The application will be available at http://localhost:8000

Using Docker

To run the application using Docker:

1. Build the Docker image:

   docker build -t chatsbt .
   

2. Run the container:

   docker run -p 8000:8000 chatsbt
   

The application will be available at http://localhost:8000

API Endpoints

The backend exposes the following RESTful API endpoints:

Models

• GET /api/models - Retrieve list of available AI models
  • Response: {"models": ["model1", "model2", ...]}

Chats

• POST /api/chats - Create a new chat session

  • Request: {"model": "model_name"}
  • Response: {"id": "chat_id", "model": "model_name"}

• GET /api/chats/{chat_id} - Retrieve chat history

  • Response: {"messages": [{"role": "human|assistant", "content": "message_text"}, ...]}

Messages

• POST /api/chats/{chat_id}/messages - Send a message to a chat

  • Request: {"message": "user_message", "model": "model_name"}
  • Response: {"status": "queued", "message_id": "message_id"}

• GET /api/chats/{chat_id}/stream?message_id={message_id} - Stream AI response

  • Server-Sent Events (SSE) endpoint that streams the AI response token by token
  • Events:
    • data: token_content - Individual tokens from the AI response
    • event: done - Indicates the response is complete