Inspiration
Based on recent personal events, we found that LinkedIn functionalities might not be the most effective tools for filtering candidates when looking for software developers. We thought GitHub would be more reliable for that, and we decided to build up a candidate-finding system implementing neural networks.
What it does
Using the GitHub API and starting from an initial profile, it shows you similar profiles and allows you to determine whether you would hire them or not. Using a neural network, it will learn from your responses and recommend more suitable candidates.
How we built it
For creating the neural network, several research had to be made, since our purpose was to gather some skills more than using an existing solution. We also created the Android app, following the Tinder style. Neural network, server and API were developed with Python. For deployment, we used Heroku.
Challenges we ran into
Creating a neural network from scratch involves some complex calculus that we had to learn before getting hands-on.
Accomplishments that we're proud of
Building a functional and fast neural net.
What we learned
Machine learning theory, Android...
Built With
- android
- github-api
- machine-learning
- mongodb
- neural-network
- numpy
- python
Log in or sign up for Devpost to join the conversation.