🦜🕸️ JARVIS - A Personalized Multi-Agent AI System

JARVIS is a sophisticated multi-agent system built on the LangGraph framework. It goes beyond a simple chatbot, providing a suite of AI agents that collaborate to perform a variety of intelligent tasks based on your local data. With a user-friendly startup script, JARVIS makes complex setup processes effortless, allowing you to harness the power of personalized AI assistants.

🌟 Core Features

• Multi-Agent Architecture: Powered by LangGraph, JARVIS orchestrates multiple specialized AI agents (Chatbot, Coding, Dashboard, Recommendation) that work together to handle complex tasks.
• Personalized RAG: The system builds a Retrieval-Augmented Generation (RAG) pipeline based on your local files. You select the folders, and JARVIS uses them as a knowledge base to provide contextually accurate answers.
• Automated Environment Setup: A single start.py script checks for dependencies (Docker, Python packages), configures the environment, and launches all necessary services.
• User-Friendly Interface: Features a desktop floating chat widget for easy access and a GUI-based setup process for user surveys and folder selection.
• Data-Driven Intelligence: Automatically collects and indexes data from your selected files, enabling agents to provide personalized insights, visualizations, and recommendations.

🛠️ System Architecture

JARVIS is composed of a FastAPI backend, a user-facing frontend, and a suite of intelligent agents.

1. Backend (/backend)

• API Server (main.py): A FastAPI server that manages agent lifecycle and handles all API requests.
• Agent Core (/agents): Contains the logic for each specialized agent:
  • ChatbotAgent: Interacts with the user via a RAG-based conversational model.
  • CodingAgent: Assists with code generation, debugging, and other development tasks.
  • DashboardAgent: Analyzes data and generates insightful visualizations.
  • RecommendationAgent: Provides personalized recommendations based on user data.
• Database (/database): Manages data collection from local files, user profiles, and metadata storage using SQLite and a Qdrant vector database.

2. Frontend (/frontend)

• Floating Chat App (front.py): A desktop widget that serves as the primary user interface for interacting with JARVIS.
• Setup Dialogs (survey_dialog.py, folder_selector.py): GUI components that guide the user through the initial setup process.

3. Launcher (start.py)

This script is the main entry point for the entire system. It performs the following steps in order:

1. Checks if Docker and required Python packages are installed.
2. Verifies the existence of the .env configuration file.
3. Starts the Qdrant vector database using Docker.
4. Initializes the SQLite database.
5. Launches the backend FastAPI server.
6. Prompts the user with a survey and a folder selection dialog.
7. Initiates the initial data collection and indexing process from the selected folders.
8. Starts the desktop floating chat application.

🚀 Getting Started

1. Prerequisites:

   • Python 3.10+
   • Docker Desktop (running)

2. Installation:

   # Clone the repository
   git clone https://github.com/your-repo/JARVIS.git
   cd JARVIS
   
   # Install dependencies
   pip install -r requirements.txt

3. Configuration:

   • You may need to create a .env file based on the startup script's guidance.
   • Add your GEMINI_API_KEY to the .env file if required.

4. Running the System:

   python start.py

   The script will guide you through the rest of the setup process. Enjoy your personalized AI assistant!