AgentVerse

Project Goal 🎯

This project is an AI Agent-powered Streamlit web application designed to assist users with queries related to the Marvel and DC comic universes.
At its core is a master agent that intelligently routes user questions to the appropriate domain-specific agent, either Marvel or DC, using a Gemini-based LLM classifier.

Users can interact with the system in two modes:

1. LLM-based assistant for dynamic reasoning and conversation
2. Dataset-driven mode where queries are answered using pre-curated information from structured Marvel and DC datasets.

The system employs modular agents to separate concerns cleanly.
Gemini API integration ensures lightweight, fast response generation.
This architecture allows scalable, accurate, and context-aware responses across both comic book domains.

Project Structure 📁

├── Agents/                  # Contains agent logic (e.g., master_agent.py, marvel_agent.py, dc_agent.py)
├── Dataset/                 # Contains Marvel/DC datasets                
├── streamlit_app.py         # Entry point for the Streamlit web application
├── utils.py                 # Shared utilities (e.g., LLM calls, search links)
├── .gitignore            
├── LICENSE                 
└── README.md 

Video Output 🎥

Watch the project demo here:

demo.mp4

Requirements 📦

To run this project, ensure you have the following dependencies installed:

• streamlit
• google-generativeai
• pandas
• python-dotenv

You can install the required packages using pip:

pip install streamlit google-generativeai python-dotenv pandas

Usage 🚀

Clone the repository:

git clone https://github.com/mrinmoycyber/AgentVerse.git

Navigate to the project directory:

cd AgentVerse

Create a .env file and add your Gemini API key:

GEMINI_API_KEY=your_api_key_here

Run the Streamlit app:

streamlit_app.py