DevBuddyAI

Inspiration

We were inspired by the hackathon category “Best Developer Tool,” which challenges participants to improve the developer experience in some meaningful way. The focus on enhancing productivity, simplifying the development process, and making coding more efficient resonated with our team. As developers ourselves, we often struggle with improving code quality, managing feedback, and optimizing workflows, so we built DevBuddyAI to tackle these common pain points by providing real-time code analysis, suggestions, and feedback powered by AI. Our goal is to make developers’ lives easier at every stage of the software development lifecycle.

What it does

DevBuddyAI is an intelligent code assistant that analyzes a developer’s code in real time and provides suggestions for improvement, comments on potential issues, and feedback on how to enhance code quality. It uses Google Gemini to read the code and generate feedback, making the development process smoother and more intuitive. Whether it’s identifying areas for optimization, offering refactoring suggestions, or flagging potential bugs, DevBuddyAI helps developers focus on writing better code faster.

How we built it

We built DevBuddyAI as a full-stack application using NodeJs and NextJS for the frontend and Flask and JavaScript for the backend. For the AI-driven code analysis, we integrated Google Gemini, which reads the code and generates feedback in real time. While our initial plan was to use BreadBoardAI, we encountered documentation and downtime issues that led us to pivot to Google Gemini for its reliability. The project was also structured to handle communication between the frontend and backend through APIs, but we faced integration challenges due to multiple ports and API design, which we eventually overcame.

Challenges we ran into

• BreadBoardAI Issues: Initially, we planned to use BreadBoardAI, but its poor documentation and unexpected downtime hindered progress. This led us to switch to Google Gemini after feedback from the judges.
• Frontend-Backend Integration: We had issues integrating the frontend and backend, particularly due to running them on separate ports and difficulties with API connection. We realized late in the project that we should have designed the REST API differently for smoother communication.
• Planning Pitfalls: Our early-stage planning didn’t fully account for the complexity of integrating AI and REST APIs, which led to delays as we had to adjust the architecture mid-project.

Accomplishments that we're proud of

• We successfully built a full-stack application from scratch and got it to compile and run despite several integration hurdles.
• We learned and worked with new technologies like BreadBoardAI and Google Gemini, gaining valuable insights into AI-powered development tools.
• We were able to adjust and pivot effectively during the hackathon, turning technical challenges into learning opportunities.

What we learned

This project taught us the importance of detailed planning before diving into implementation, especially when working with AI and APIs. We learned to account for potential integration issues upfront and to choose technology stacks that can scale with our needs. Additionally, we gained a deeper understanding of how to use AI tools to enhance the developer experience and how to manage cross-functional communication between frontend and backend services.

What's next for DevBuddyAI

We are excited to expand DevBuddyAI further by:

• Adding an attachment feature that allows users to upload files directly for AI analysis.
• Developing a VSCode plugin, which was our initial vision for the project, to seamlessly integrate DevBuddyAI into developers’ workflows, making it even easier to use and more accessible to developers worldwide.

Built With

• axios
• flask
• google-gemini
• insomnia
• javascript
• next.js
• node.js
• python
• react.js
• vscode