CodeSense AI — AI Code Reviewer & Error Detector

An AI-powered web application that reviews source code, detects bugs, explains problems, and suggests improvements. Built with Django, Nvidia NIM (Llama 3.1 70B), and Judge0 for remote code execution. Ready for Serverless deployment on Vercel.

Features

• 🔍 AI Code Review — Get structured feedback on bugs, security, performance & style using Llama 3.1 70B.
• 💬 AI Chat — Ask follow-up questions about your code and reviews.
• ⚡ Live Code Execution — Run code directly in the browser (Python, C++, Java, JS, etc.) via Judge0 API.
• 🪟 LeetCode Style UI — Fully resizable, split-pane layout with independent scrolling and Monaco Editor.
• 🌙 Dark/Light Theme — Developer-friendly modern interface.
• ☁️ Vercel Ready — Designed to be deployed effortlessly on Vercel as Serverless Functions.

Tech Stack

• Backend: Django 6.0, Python
• Frontend: Vanilla JS, CSS (Stitch Design System), Monaco Editor
• AI Inference: Nvidia NIM API (meta/llama-3.1-70b-instruct)
• Code Execution: Judge0 CE API

Quick Start (Local Setup)

Prerequisites

• Python 3.10+
• Free Nvidia API Key from build.nvidia.com

1. Clone & Install

git clone https://github.com/shubhamjrd4559-sudo/CodeSenseAI.git
cd CodeSenseAI

# Create virtual environment (recommended)
python -m venv venv
source venv/bin/activate  # Linux/Mac
# venv\Scripts\activate   # Windows

# Install dependencies
pip install -r requirements.txt

2. Configure Environment

Create a .env file in the root directory and add your Nvidia API Key:

NVIDIA_API_KEY="nvapi-your-key-here"
LLM_MODEL="meta/llama-3.1-70b-instruct"

3. Run Django

python manage.py runserver

Open http://localhost:8000 in your browser.

Deployment (Vercel)

This project is pre-configured with vercel.json and wsgi.py optimizations for 1-click serverless deployment.

1. Push your code to a GitHub repository.
2. Sign in to Vercel and click Add New Project.
3. Import your repository.
4. In the Environment Variables section, add:
   • NVIDIA_API_KEY: Your Nvidia API key
   • LLM_MODEL: meta/llama-3.1-70b-instruct
5. Click Deploy.

Usage

1. Write/paste code in the right-side editor panel.
2. Select language and review mode from the top header.
3. Click Review (or Ctrl+Enter) to get AI analysis on the left panel.
4. Click Run (or F5) to execute the code and view the output in the bottom terminal.
5. Drag the borders between panels to resize the layout to your preference.

License

MIT