pareview

AI code reviews that don't just fix your code—they teach you how to become a better programmer.

Comment

Inspiration

Learning to code can be challenging, especially when beginners only receive error messages or code fixes without understanding why the issue happened. We wanted to create an AI-powered code reviewer that not only finds bugs but also teaches the programming concepts behind each suggestion. That idea became Pareview—an AI code reviewer with a teaching layer that helps developers learn while improving their code.

What it does

Pareview allows users to paste JavaScript, Python, or HTML/CSS code into a simple web interface. The application sends the code to the Qwen Cloud API, which reviews it for bugs, logic errors, security risks, coding style issues, and best practices.

Unlike traditional code reviewers, Pareview explains every issue in plain English. Each review includes:

The detected problem Why it matters A suggested fix A short educational explanation so the user understands the concept and avoids making the same mistake again. How we built it

Pareview was built using HTML, CSS, and JavaScript for the frontend. AI-powered code reviews are generated using the Qwen Cloud API. The project is hosted online, with the source code managed on GitHub.

During development, AI tools including ChatGPT and Claude were used for brainstorming, debugging, improving the user interface, refining documentation, and testing ideas while Qwen Cloud powers the application's core AI review functionality.

Challenges we ran into

The biggest challenge was integrating the AI review system while keeping the interface simple and easy to understand. Another challenge was designing responses that were educational instead of simply returning code fixes. We also learned how to structure prompts for better AI feedback and prepared deployment materials required for the hackathon.

Accomplishments that we're proud of Built a working AI-powered code reviewer. Added a teaching layer that explains programming concepts instead of only providing fixes. Supported multiple programming languages including JavaScript, Python, and HTML/CSS. Successfully integrated the Qwen Cloud API into the application. Created a clean and beginner-friendly interface suitable for students and new developers. What we learned

This project helped us learn how to integrate large language models into a web application, design better prompts for code analysis, improve frontend development skills, and understand the importance of making AI educational rather than simply automated. We also gained valuable experience preparing a complete hackathon submission, including documentation, architecture diagrams, deployment, and project presentation.

What's next for Pareview

Our future plans include:

Supporting more programming languages such as C++, Java, and C#. Providing line-by-line code annotations. Adding AI-generated code quality scores. Allowing users to compare versions of their code. Integrating Pareview directly into IDEs such as Visual Studio Code. Creating personalized learning recommendations based on common coding mistakes.

Built With

Share this project:

Updates

Submission history