MentorZero is a Streamlit-based chatbot that channels the persona of a YC partner to give direct, practical, and context-grounded advice to founders. It features persistent chat history, dynamic chat UI, and supports both RAG (retrieval-augmented generation) and OpenAI completions.

• Dynamic, modern chat UI (WhatsApp/ChatGPT style)
• Persistent chat history and conversation management
• Source attribution for retrieved knowledge
• Displays token usage, API cost, and response time
• Light and dark mode support (customizable)
• Easily deployable to Replit, Streamlit Cloud, or locally

• Python 3.8+
• Streamlit
• OpenAI Python SDK

Install dependencies:

pip install -r requirements.txt

1. Add your OpenAI API key:
   • Set the OPENAI_API_KEY environment variable, or edit chatbot_app.py to set it directly.
2. Run the app:

   streamlit run chatbot_app.py

3. Open your browser:
   • Go to http://localhost:8501 (or the port shown in your terminal)

• Push this repo to GitHub
• Import into Replit
• Set the run command to:

  streamlit run chatbot_app.py --server.port=8080 --server.address=0.0.0.0

• Add your OpenAI API key as a secret in Replit

.
├── chatbot_app.py           # Main Streamlit app
├── chatbot_style.css        # Custom chat UI styles (dark)
├── chatbot_style_light.css  # Custom chat UI styles (light)
├── rag_logic.py             # Retrieval-augmented generation logic
├── requirements.txt         # Python dependencies
├── README.md                # This file
└── ...                      # Data and other modules

This project is for educational and prototyping purposes. Please check OpenAI's terms for API usage.

Built by Samarjit, powered by Streamlit and OpenAI.