OLiRE

Inspiration

Developing software often comes with tight deadlines. Shifting thorough large pile of internet data is often time-consuming dead end. Devs needed a tool that could quickly surface the latest, most relevant libraries that perfectly fit their project's needs. This frustration was what led OLiRE - a recommendation engine that cuts through the noise and helps developers find the perfect open-source tools, fast.

What it does

Tell it your project, and OLiRE recommends you the best open-source libraries based on your needs. Saves you time and frustration. Think of it as matchmaker for code. Open-source, free, and developer-friendly.

How we built it

The frond-end and the API endpoints use Next.Js 14. Behind the scenes, a python crawler will search the web for new libraries, extract and clean the data and then store it in a MySQL server. The python crawler is dockerized, so that when the need comes for scalability it will be as easy as spinning up a new container.

Challenges we ran into

Dockerizing Next.Js was a little challenging because Node-18 Alpine image had some networking issue and it was poorly documented. Additionally, the internet's unstructured nature made fully automated data collection a beast. I opt for a semi-automated approach by manually curating the initial data needed for an MVP.

Accomplishments that we're proud of

• Built a comprehensive knowledge base of open-source libraries and frameworks
• Successfully dockerized the front-end and back-end components, for easy deployment and scalability.
• Devised a semi-automated data collection approach to address the unstructured nature of web data. ## What we learned This project was a fantastic learning experience. I learned about exploring strategies for creating and maintain a comprehensive database of open-source libraries. Gained an admiration and hand-on experience with dockerizing applications.

What's next for Olire

Integration with developer communities like Reddit, Hackernews and more are on the horizon. Additionally, I planned to incorporate filtering techniques for better recommendation based on user data. While this was deemed too ambitious for an initial MVP, I think it holds immense potential for enhancing the accuracy and personalization of OLiRE's recommendation.

Built With

• bash
• docker
• mysql
• nextjs
• phpmyadmin
• python
• react
• sql
• tailwind
• typescript