🤖 MongoChat AI Platform

MongoChat is an intelligent, conversational interface for your MongoDB database. It allows you to chat with your data using natural language, powered by advanced RAG (Retrieval-Augmented Generation) and Large Language Models (LLMs).

Forget writing complex aggregation pipelines manually—just ask, and MongoChat understands the structure of your data and helps you retrieve it.

---

Dashboard Chat Agent

Convertion - Chat

🚀 Key Features

• ⚡ Agentic Data Retrieval: The AI simply doesn't just "talk"—it ACTS. If you ask "Count the users", it securely executes the query and gives you the real number.
• 🧑‍💻 Smart Code Generation: If you ask "How do I write a query?", it switches modes to teach you, providing clean, copy-pasteable code without internal tool jargon.
• 🗣️ Natural Language Chat: Ask complex questions in plain English.
• 🔌 Secure Connection: Connect to any MongoDB cluster (Atlas or local) securely. Credentials are ephemeral and stored only in your session.
• 🧠 RAG Engine: The system scans your database schema, indexes it, and uses this knowledge to give accurate answers.
• 📝 Comprehensive Logging:
  • Application Logs: Detailed system events in logs/mongo_chat.log.
  • Audit History: Tracks who asked what, including System IPv4 (LAN IP) logging for local users.
• ⏱️ Real-Time UX: AJAX-based chat, auto-scrolling, and per-message timestamps.

---

📂 Project Structure

Here is a map of the files in this project:

mongo_chat_platform/            # 📁 ROOT DIRECTORY
│
├── manage.py                   # ⚙️ Django's command-line utility
├── run.sh / run.bat            # 🚀 Scripts to start the server (Auto-cleans caches)
├── requirements.txt            # 📦 List of python libraries required
├── README.md                   # 📖 This documentation file
├── .env                        # 🔐 Secrets (API Keys) - Create this yourself!
│
├── mongo_chat_platform/        # 🧠 CORE CONFIGURATION & SERVICES
│   ├── settings.py             # Global settings (Installed apps, Middleware)
│   ├── logger.py               # Central logging setup (Console + File)
│   └── services/               # The "Brain" of the application
│       ├── mongo_service.py    # Connects to your MongoDB to execute queries (Agentic)
│       ├── chroma_service.py   # Vector DB (RAG) to remember your schema
│       ├── llm_service.py      # Talks to Groq AI (Llama 3)
│       └── logging_service.py  # Saves chat history to a database
│
├── connect/                    # 🔌 APP: CONNECTION SCREEN
│   ├── views.py                # Handles login/logout logic (AJAX enabled)
│   └── templates/connect/      # HTML for the Connection Form
│
├── chat/                       # 💬 APP: CHAT INTERFACE
│   ├── views.py                # Handles messaging, tool execution loop & logging
│   └── templates/chat/         # HTML for the Chat Interface (Timestamps enabled)
│
├── templates/                  # 🎨 SHARED TEMPLATES
│   └── base.html               # Main layout file (includes Tailwind CSS)
│
├── logs/                       # 📝 LOG FILES
│   └── mongo_chat.log          # Detailed application logs appear here
│
└── .husky/                     # 🛡️ GIT HOOKS
    └── pre-commit              # Script that blocks you from committing .env

---

⚙️ Setup Guide

Prerequisites

• Python 3.10+ (Ensure it is added to your PATH)
• MongoDB Database (Local or Atlas Connection String)
• Groq API Key (Free beta keys available at console.groq.com)

1. Installation

Clone the repository and navigate to the folder:

git clone https://github.com/yourusername/mongo-chat-platform.git
cd mongo-chat-platform

Install the required Python packages:

pip install -r requirements.txt

2. Configuration

Create a .env file in the mongo_chat_platform folder (next to settings.py) with the following keys:

# Security
SECRET_KEY=your_django_secret_key
DEBUG=True

# LLM Provider
GROQ_API_KEY=gsk_your_groq_api_key_here
GROQ_MODEL=llama-3.3-70b-versatile

# Logging (Optional: Where to store chat logs)
MONGO_LOGS_URI=mongodb://localhost:27017/chat_logs

# ChromaDB (Optional: For cloud vector storage)
# CHROMA_API_KEY=...

3. Running the App

Run the database migrations and start the server:

Windows: Double-click run.bat or run:

.\run.bat

Linux/Mac:

./run.sh

Visit http://127.0.0.1:8000 in your browser.

---

📖 How to Use

1. Connect: Enter your MongoDB Connection String (e.g., mongodb+srv://user:pass@cluster...) on the home screen.
2. Wait: On the very first run, the system will download a small AI embedding model. This takes about 1-2 minutes.
3. Chat: Once connected, you'll be taken to the chat interface.

Agentic Capabilities (New!)

• Ask for Data: "How many users are active?"
  • The AI performs a live count in the DB and answers: "There are 150 active users."
• Ask for Code: "Show me the query to find active users."
  • The AI switches mode and gives you the clean code block: db.users.find({...}).

---

🛡️ Security & Contribution

• Credential Safety: We never store your MongoDB URI in a persistent database. It lives only in your temporary session.
• Git Protection: A husky pre-commit hook is installed to prevent you from accidentally committing your .env file.
• Logging: All interactions are logged with timestamps and System IP (even for localhost users) to ensure accountability.

Feel free to fork this project and submit Pull Requests!