Inspiration
Being an engineering student , long documentation seems to heavy or long for me and during reading documentation if i got stuck somewhere then to understand that particular point i took help by opening new tab (chatgpt , gemini) .So , I decided to build AI-powered code understanding right where developers /new engineering student need it — inside their documentation.
What it does
DevDocs+ contains three buttons - 1. *Explain button *- If something seems difficult to me and could not able to understand by reading documentation then i will select that particular part and ask button appear after clicking on it it automatically open DevDocs+ window. 2.Summarise button - It provides 2 -4 lines of any term/concept to get overview with code example. 3. Search button - This is the most special button sometimes we as engineer want to add some feature but do not know its technical term so i can write prompt in general language then it will provide link for that particular feature/code.
How we built it
Frontend:
1.Chrome Extension built using JavaScript, HTML, and CSS 2.Injects a floating “Ask DevDocs+” button when text is selected 3.Sends API requests to the backend and displays formatted AI responses
Backend:
1.Built with FastAPI 2.Integrated with Groq API for intelligent code explanations and summaries 3.Uses CORS middleware to handle extension requests 4.Structured routes for modular AI endpoints
Tech Stack: 🧩 FastAPI (Python) ⚙️ Groq AI API 💻 JavaScript (Chrome Extension) 🌐 HTML + CSS
Challenges we ran into
1.API Rate Limits: Groq API has strict quota limits (HTTP 429 errors), so we implemented request throttling and retries. 2.CORS Issues: The browser extension couldn’t directly communicate with the backend initially. We solved this using FastAPI’s CORS middleware. 3.Formatting AI Responses: The responses from the API were often dense or unformatted. We used Markdown rendering and whitespace optimization to make it visually clear
Accomplishments that we're proud of
I made this project unique as i analyse real world problem and try to solve it in easiest way that help engineers like me
What we learned
I learned many things like how to connect python backend to html/css/js frontend and deploying project on render and managing project with my daily studies .
What's next for DevDocs+
1.Adding multilingual support for documentation. 2.Integrating with VS Code for in-editor explanations. 3.Improving output more like for developer style
Built With
- css
- fastapi
- groq
- html
- javascript
- python
- render
Log in or sign up for Devpost to join the conversation.