Hk669 / Open-Source-Recommender

Gitmatch - personalized recommendation system, that recommends open-source repositories based on your skills, topics and languages from your GitHub
https://gitmatch.in/
MIT License
4 stars 0 forks source link
ai aws azure-openai chromadb fastapi git-search gitmatch good-first-issue open-source-recommender recommendation-system redis-cache repository-recommender

Open-Source Recommender

https://github.com/user-attachments/assets/fc2f9fbe-682e-406d-8c7b-d9f192889a4f

This will be a free public-facing web application designed to find open-source projects for beginners and developers available at gitmatch.in.

Search your next contribution to open source easily! A free web app is here to help every developer find cool open-source projects of interest that fit their skillset. Just enter your GitHub username; our intelligent recommender system will do the rest.

Give it a star⭐, if you support my intiative to help beginners start with open source!

Why Use Open-Source Recommender

How It Works

architecture

  1. Retrives user repositories details which include, languages, topics, and description
  2. Collects the best open source projects from the GitHub based on user's topics and languages
  3. Processes the open source repositories through the embedding model deployed in the Azure OpenAI Studio
  4. Stores the resulting embeddings in ChromaDB (VectorStore)
  5. Converts the user's repository languages, topics, and descriptions into embeddings using the embedding model
  6. Perform a similarity search with the embeddings to find the most relevant open source projects
  7. Delivers personalized recommendations to the client

Features

feel free to drop your suggestions and issues at Discussions

Get Started

checkout gitmatch.in, and if you want to leverage the true power of the gitmatch, follow steps

Feedback

If you have any suggestions, find any bugs, or have success stories you'd like to share with me, please do so. Your input will make the experience better for all.

License

This project is licensed under the MIT License. See the LICENSE file for details.