The phase consists of getting access to possible data source(s), pulling data on regular basis, analysis of dataset content and format.
Having analyzed the data in a proper way we would be able to say what we can predict, suggest how to visualize it in a best way.
Integration with Cleeng server through authentication and API calls, development and exposing API to retrieve prediction result from churn prediction platform, developing API documentation.
Model selection upon available facts after feature engineering.
Creation of services to handle main set of methods like: create, post, put, delete.
Feature engineering is the process of using domain knowledge of the data to create features that make machine learning algorithms work.
Front-end web development using AngularJS 1.x to present prediction results.
Platform architecture and technology stack selection based on selected model and available data sets. Setup AWS instances and enabling essential processing/storage services.
Calculating essential model parameters and verifying model reliability (supervised machine learning).
The phase includes: testing of prediction model, development part (back-end and front-end functionality).