AWS Builder Center
Weekend Creative Challenge: Roast My Code

Weekend Creative Challenge: Roast My Code

Roast My Code is an AI-powered developer tool that turns code reviews into a fun experience. Paste your code, choose a programming language and roast intensity, and get a humorous roast along with technical issues, severity ratings, improvement suggestions, and cleaner refactored code. Designed with an 8-bit retro arcade aesthetic, Roast My Code combines developer humor, creativity, and practical coding feedback in one interactive experience.

AWS Student Builder Leader | Full-Stack Developer | Cloud Enthusiast

Vision & What the App Does

Code reviews do not always have to be serious. Roast My Code is a developer-focused creative application that turns the process of reviewing code into an entertaining experience. Its tagline is simple: "Your code has bugs. We have jokes."
The idea behind the application is to combine developer humor with useful technical feedback. Users paste their source code, select a programming language, and choose a roast intensity: Gentle, Savage, or Nuclear. The application then analyzes the code and produces a humorous roast alongside a technical explanation of the problems, severity ratings, improvement suggestions, and a cleaner refactored version of the code.
The app supports 14 programming languages, including JavaScript, TypeScript, Python, Java, C, C++, C#, PHP, Go, Rust, HTML, CSS, SQL, and Other. Users can also copy the improved code or create a shareable PNG card containing their roast, severity level, and developer verdict.

How I Built It

The application was built using React 19, TypeScript, Vite, and Tailwind CSS. The interface was designed around an 80s retro arcade and cyberpunk terminal aesthetic, combining pixel-art visuals, neon colors, chunky borders, and arcade-style typography.  Press Start 2P  is used for the arcade-inspired interface, while JetBrains Mono and monospace fonts are used for the coding experience.
The main interaction is intentionally simple: paste code, choose the language and roast level, and click the ROAST button. While the application processes the request, a retro loading animation displays humorous messages. The results then show the roast, technical issues, severity score, improvement suggestions, and refactored code. A confetti effect celebrates successful results.
One of the main technical challenges was handling requests through the serverless backend. Browser requests initially encountered CORS preflight issues, which required explicit handling of  OPTIONS  requests and HTTP methods. Another issue involved Base64-encoded request bodies, which was solved by decoding the payload before parsing the JSON.
I also implemented multi-model failover so the application can continue processing when one AI model encounters temporary availability or quota issues. Another challenge was generating downloadable share cards while preserving the application's custom pixel-art styling, shadows, SVG elements, and fonts. This was solved using client-side image rendering with  html-to-image .

AWS Services & Architecture

The application uses AWS Amplify for frontend hosting and automated deployment from the GitHub repository. AWS Lambda provides the serverless backend, while a Lambda Function URL provides the HTTPS endpoint used by the frontend. AWS IAM is used to manage permissions for the Lambda function.
The architecture follows a lightweight serverless approach:
User Browser
↓
AWS Amplify Hosted React App
↓
AWS Lambda Function URL
↓
AI Processing
↓
Structured Response
↓
React Application
↓
Roast + Analysis + Refactored Code
The backend receives the submitted code, programming language, and selected roast intensity, processes the request, and returns structured information containing the roast, severity score, technical summary, issues, improvements, better code, and developer verdict.

What I Learned

Building Roast My Code provided practical experience with serverless application development and generative AI integration. One of the biggest lessons was learning how to connect a modern frontend application to a lightweight serverless backend without maintaining a traditional server.
I also learned the importance of structured AI responses. Instead of displaying a single block of generated text, the application processes structured information so that the roast, technical issues, improvements, and refactored code can each be displayed in their own UI components.
Another important lesson was building a more resilient application by handling CORS issues, encoded request payloads, model availability, and client-side image generation.
Finally, the project showed that developer tools do not have to look boring. Combining an 8-bit retro arcade aesthetic with modern AI created a tool that is both useful and entertaining.

Project Link

GitHub Repository:  Roast My Code on GitHub 
Roast My Code combines creative expression, developer humor, AI, and AWS serverless technology into a simple experience designed to make code reviews a little more fun.
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article