Skip to content

Latest commit

 

History

History
132 lines (104 loc) · 4.33 KB

README.md

File metadata and controls

132 lines (104 loc) · 4.33 KB

RAG Chat - AI Assistant for My Portfolio

A Retrieval-Augmented Generation (RAG) chatbot that answers questions about my professional background and portfolio using Claude 3.5 Haiku.

RAG Chat Screenshot

Features

  • AI-Powered Responses: Uses Anthropic's Claude 3.5 Haiku to generate natural, informative responses
  • Context-Aware: Retrieves relevant information from my portfolio content
  • Rate Limiting: Prevents abuse with client and server-side rate limiting
  • Responsive Design: Works on desktop and mobile devices
  • References: Provides links to relevant sections of my portfolio
  • Analytics: Tracks usage patterns with Vercel Analytics

Technology Stack

  • Frontend: HTML, CSS, JavaScript
  • Backend: Python with FastAPI
  • AI: Anthropic Claude 3.5 Haiku for text generation
  • Embeddings: OpenAI's text-embedding-ada-002 for semantic search
  • Deployment: Vercel Serverless Functions
  • Analytics: Vercel Analytics

How It Works

  1. Embedding Generation: Content from my portfolio is converted into vector embeddings
  2. Query Processing: User questions are converted to embeddings and matched with relevant content
  3. Context Building: Matched content is used to create context for the AI
  4. Response Generation: Claude generates a response based on the context and question
  5. Reference Linking: Links to relevant portfolio sections are provided

Local Development

Prerequisites

  • Python 3.9+
  • Node.js (for frontend development)
  • OpenAI API key
  • Anthropic API key

Setup

  1. Clone the repository:

    git clone https://github.com/sohae-kim/rag-chat-resume.git
cd rag-chat

  2. Create a virtual environment:

    python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

  3. Install dependencies:

    pip install -r requirements.txt

  4. Create a .env file with your API keys:

    OPENAI_API_KEY=your_openai_key
ANTHROPIC_API_KEY=your_anthropic_key

  5. Generate embeddings:

    python scripts/build.py

  6. Run the development server:

    python run.py

  7. Open http://localhost:8000 in your browser

Deployment

This project is configured for deployment on Vercel as a serverless function:

  1. Connect your GitHub repository to Vercel
  2. Set environment variables in the Vercel dashboard:
    • OPENAI_API_KEY
    • ANTHROPIC_API_KEY
  3. Deploy!

Project Structure

rag-chat-resume/
├── api/                    # Vercel serverless functions
│   ├── index.py            # Main API endpoint
│   ├── chat.py             # Chat handler for AWS Lambda/Vercel
│   └── __init__.py         # Package initialization
├── app/                    # Backend application
│   ├── main.py             # FastAPI application
│   ├── utils.py            # Utility functions
│   ├── logging.py          # Logging configuration
│   └── __init__.py         # Package initialization
├── data/
│   ├── content.json        # Portfolio content
│   └── embeddings.json     # Generated embeddings (gitignored)
├── public/                 # Static files
│   ├── static/
│   │   ├── js/             # JavaScript files
│   │   ├── css/            # CSS files
│   │   └── images/         # Images
│   └── index.html          # Main HTML file
├── scripts/
│   ├── build.py            # Script to generate embeddings
│   └── generate_embeddings.py # Alternative embeddings generator
├── .env                    # Environment variables (gitignored)
├── .gitignore              # Git ignore file
├── next.config.js          # Next.js configuration for Vercel
├── package.json            # Node.js package configuration
├── requirements.txt        # Python dependencies
├── vercel.json             # Vercel configuration
└── README.md               # This file

License

MIT

Acknowledgements