CodeFuse

Logo
Banner
CodeFuse
Code Explainer
Calculating the time complexity of a program
Code Language Translator
SQL Translate

Inspiration

We are huge fans of how Artificial Intelligence can transform and channel human lives to maximize efficiency and optimize outcomes. Recently, we were quite amazed by GitHub Copilot, which greatly assists developers in writing smart and efficient code. So, we thought of integrating certain AI features that could help us to use Confluence in an effective way.

Introducing "CodeFuse - A simple, yet powerful, AI-powered code helper"

What it does

Code Explainer: The first thought that came into our mind was to build something that could explain a piece of code in plain natural language directly from a Confluence page. Sounds interesting? Yes, of course, it is! Moreover, it would also help those who don’t have a deeper understanding of programming in a very efficient way to learn about the actions of any code snippet.
Calculating the time complexity of a program: It is critical for developers to check whether or not their function or program is optimized. Assume you're adding a code snippet to your Confluence page that has a time complexity of O(n³)! Wouldn't it be great if you could detect and predict the time complexity of your program on the fly using a super handy tool?
Code Language Translator: Suppose you are a JavaScript developer and you found some piece of code on the Confluence page which is written in Python. Pain, right? But wait, we are here to help you. What if you could convert a code snippet from any language to your desired language easily with just a click!
SQL Translate: Now developers who need to deal with databases (especially SQL databases like PostgreSQL and MySQL, which are widely used) also need to write complex schemas and database queries. People from some other teams can note down the requirements that developers need to structure the queries. For this purpose, we also wanted to introduce an AI-based tool that can help developers write effective SQL queries from plain text.

How we built it

We used Forge CLI to create and deploy the Forge UI of the project. The backend is based on node.js and express.js while the OpenAI API is used to give our app the power of the GPT-3 AI Model.

Challenges we ran into

Sanitizing the code snippet without the line numbers was really painful as we needed to take care of so many edge cases.

Accomplishments that we're proud of

We are proud that we were able to successfully build this project within just 3 days.

What we learned

It was a great experience for us to participate in this hackathon. We learned to use forge along with OpenAI.

What's next for CodeFuse

We have plans to add more features to make our product more effective and powerful. Also, we are planning to launch the app on the Atlassian Marketplace in the future.