Code Angel

Landing Page
Classes Page for student
Assignment Page for Students
Assignment Page
Interacting with Integrated AI helper
Instructor view of students
Instructor can see chat logs from student

Inspiration

The idea for Code Angel came from the growing need to integrate AI into coding education responsibly. We wanted to create a platform that allows students to leverage AI assistance while keeping their work auditable for instructors, essentially creating a “paper trail” similar to how math classes allow the use of properly cited prior results.

What it does

Code Angel is an AI-assisted coding tool that provides students with intelligent support while keeping instructors in the loop. Students can interact with AI to generate code, get hints, or debug programs, and every interaction is logged in a way that instructors can review. This allows teachers to monitor AI usage, ensure proper attribution, and maintain academic integrity in coding assignments.

How we built it

We built Code Angel using Next.js for the frontend, Supabase for authentication and database management, and AWS Bedrock to power the AI backend. Next.js allowed us to build a responsive, interactive interface, Supabase handled user accounts and session management, and Bedrock enabled us to integrate state-of-the-art AI models capable of understanding and generating code. The combination gave us a full-stack solution that is both scalable and secure.

Challenges we ran into

Balancing usability and auditability: We needed to make the AI helpful for students while ensuring all interactions were logged transparently for instructors.
Integrating AI with real-time logging: Capturing every AI query and response in a structured way without slowing down the student workflow was tricky.
Handling edge cases in code generation: Sometimes the AI would produce incorrect or unsafe code. We had to implement safeguards and guidance prompts.
Authentication and roles: Ensuring that students and instructors had the correct permissions and access levels required careful Supabase configuration.

Accomplishments that we're proud of

Successfully implemented a full-stack AI-assisted coding platform that supports instructor auditing.
Built a seamless integration between Next.js, Supabase, and AWS Bedrock.
Created a real-time logging system that captures AI interactions without interrupting the student workflow.
Developed a user-friendly interface that students can easily navigate while still giving instructors full oversight.

What we learned

How to architect a full-stack application combining modern frameworks and cloud AI services.
The importance of logging and audit trails in educational technology.
How to handle authentication, role-based access, and secure database interactions with Supabase.
Strategies for making AI tools helpful without enabling misuse, a critical balance in academic contexts.

What's next for CodeAngel

Expand AI capabilities to provide more contextual hints and debugging assistance.
Improve analytics for instructors to better understand student engagement and AI usage patterns.
Introduce collaboration features so multiple students can work together with AI guidance while maintaining accountability.
Explore integrations with popular learning management systems to make adoption easier for schools and universities.