Byte of Me

A full-stack portfolio with a free-tier AI assistant powered by a RAG pipeline. Explore my projects and skills through natural language queries, with support for i18n, dark/light mode, and responsive design.

Byte of Me is a full-stack developer portfolio that highlights my projects, skills, and experience. It includes an AI assistant built on a Retrieval-Augmented Generation (RAG) pipeline using **LangGraph **, part of the LangChain ecosystem.

The system supports multiple language models and embedding providers.

RAG Pipeline

• Gemini 2.0 Flash
• Gemini 2.0 Flash Lite
• Gemini 2.5 Flash

• text-embedding-004 - optimized for English-language tasks
• Jina Embedding v3 - designed for strong multilingual performance

• Pinecone for semantic search

• Cohere Rerank 3.5

• PostgresSaver (Supabase PostgreSQL) for storing conversation threads and chat history

The assistant provides context-aware responses in multilingual, supporting:

• Stateful AI Assistant
• Internationalization
• Multilingual UI with i18n
• Dark/Light theme toggle
• Responsive design for both mobile and desktop

Try asking: "What projects has Phu built?" or "What is Phu’s tech stack?"

1. Clone the Repository:

   git clone https://github.com/lthphuw/byte-of-me.git
   cd byte-of-me

2. Install Dependencies:

   pnpm install

3. Set Up Environment Variables:

• Create .env.local or .env files in apps/web, packages/ai, and packages/db as described in their respective READMEs:
  • Web Frontend (apps/web)
  • AI Assistant (packages/ai)
  • Database (packages/db)
  • Logger (packages/logger)

4. Run the Development Server:

• From the root directory:

  pnpm dev

• Open http://localhost:3000 in your browser.

5. Deploy to Vercel:

• Push your code to GitHub.
• Connect your repository to Vercel via the Vercel dashboard.
• Add module-specific environment variables to Vercel’s Settings > Environment Variables (see submodule READMEs).

Contributions are welcome! Please open an issue or submit a pull request to suggest improvements or report bugs.

This project is licensed under the MIT License. See LICENSE for details.