The repository contains a POC of an HR application used to predict if an employee is about to quit.
Backend is written in Django, frontend is using HTML, CSS and JavaScript (template downloaded from BBBootstrap, needed to be adjusted a bit).
Logo icon designed by Adrien Coquet.
Application is deployed on AWS Elastic Beanstalk: link
- Prediction is done based on IBM HR Analytics Employee Attrition & Performance dataset,
- EDA and model training is done in my Kaggle notebook: Employee Attrition Rate.
- To use this application, an HR worker will fill in a survey about the given employee,
- There are 26 questions in total, and they are based on the mentioned IBM HR Analytics dataset,
- Different dataset will result in different set of survey questions, it all depends on the company's data,
- HR worker using the application is acknowledged with the questions, so I didn't attach any explanations of them (please see the IBM dataset description for more information about it).
- In the ideal world, this whole process of predicting employee atrrition would be synchronized with company's HR system. Once HR has all the necessary data about the worker (for example after a quarterly interview), ML model would be triggered and provide the prediction results back to the HR system.
