BIT Code Master

An AI powered code evaluation and review system

Comment

InspirationInspiration

The idea behind BIT Code Master came from real student struggles during coding assessments. Most college platforms judge solutions only on output — ignoring logic quality, structure, and learning feedback. We wanted a system that not only checks correctness but teaches, giving personalized feedback like a mentor rather than a compiler.

What it does

BIT Code Master is an AI-powered code evaluation and learning platform where students can:

Submit code in multiple programming languages

Get real-time compilation and execution

Receive AI-generated feedback on logic, structure, best practices, and potential errors without revealing the correct solution

Track progress, accuracy, and improvement over time

Access curated coding challenges and learning materials

It turns coding practice into a guided learning experience.

How we built it

Backend: Spring Boot

Frontend: (React / Next.js / or your choice — add details)

AI Model: DeepSeek-Coder-1.3B finetuned for syntax and logic evaluation

Database: Supabase

Deployment: AWS EC2

Architecture: Students submit code → backend compiles & runs test cases → LLM analyzes structure → system generates actionable feedback → results returned to UI

The core pipeline ensures safe execution of code inside a sandbox and uses the LLM only for logic analysis (no solution leaking).

Challenges we ran into

Ensuring LLM feedback does not give away the full solution

Maintaining secure and isolated code execution for multiple users

Designing scalable architecture that doesn't overload servers during heavy compilation requests

Training the model to identify logical faults without focusing on optimization

Integrating real-time communication between compiler, model, and UI

Accomplishments that we're proud of

Built the first coding platform in our college that evaluates not just output but logic quality

Successfully deployed LLM pipeline on AWS with fast inference

Created an intuitive interface that motivates learning and skill improvement

Enabled a safe, reliable compiler environment with sandboxing

What we learned

How to integrate AI code models into real products

Deep understanding of Spring Boot microservices and API pipeline

Deployment and scaling on AWS EC2

Prompt engineering for educational feedback

Managing secure code execution environments

What's next for BIT Code Master

Support for additional languages (C++, Python, JavaScript, Kotlin)

Competitive leaderboard and coding rooms for virtual contests

Automated plagiarism detection

AI-powered hints and mini-lessons based on user errors

Mobile app for practice anywhere

Integration with college examination systems

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for BIT Code Master

Built With

Share this project:

Updates