ROCCYK_AI

ROCCYK_AI is a custom-built, open-source AI chatbot trained on the personal biography of Rhichard Koh. Designed as a personalized conversational agent, this project showcases how retrieval-augmented generation (RAG) and fine-tuned language models can be used to create highly customized chatbot experiences.

🧠 About

This chatbot serves as an AI-powered digital twin that can answer questions, hold conversations, and share insights based on the real-life biography of its creator. The project demonstrates the capabilities of personalizing large language models for niche, individual use cases, such as portfolio enhancement, education, and digital identity applications.

🚀 Features

• 🔍 Biography-trained model: The chatbot is trained on a dataset curated from the creator’s academic and professional history.
• 💬 Conversational agent: Provides natural responses in Q&A format.
• 🛠️ Modular architecture: Easy to extend or adapt for other personalized use cases.
• 📖 RAG (Retrieval-Augmented Generation): Optionally supports integration with vector stores for context-based responses.

🏗️ Tech Stack

• Python
• LangChain / Haystack (optional, for RAG)
• OpenAI GPT / LLM APIs
• FAISS / Chroma (for embedding search)
• Streamlit / Flask (optional frontend)

📁 Repository Structure

ROCCYK_AI/
├── data/               # Biographical documents
├── chatbot/            # Core chatbot logic
├── embeddings/         # Vector DB and retriever
├── ui/                 # Optional frontend interface
├── requirements.txt    # Python dependencies
└── README.md           # Project overview

🧑‍💻 Setup Instructions

1. Clone the repository

   git clone https://github.com/ROCCYK/ROCCYK_AI.git
   cd ROCCYK_AI

2. Install dependencies

   pip install -r requirements.txt

3. Prepare your .env file Include any necessary API keys (e.g., OpenAI key).

4. Run the chatbot Depending on implementation:

   python chatbot/main.py

   Or for web:

   streamlit run ui/app.py

📚 Example Use Cases

• Portfolio-based AI assistant
• Personal biography showcase
• Academic or career Q&A
• Demo project for LLM fine-tuning

🤖 Demo Questions

• “What program did Rhichard Koh graduate from?”
• “Describe Rhichard’s work on ASL translation.”
• “What AI tools has Rhichard used?”

📜 License

This project is open-source and available under the MIT License.

🙋‍♂️ Author

Built by Rhichard Koh – feel free to connect or contribute!