Hands-on Tutorial
Part 3 — End to End Machine Learning Model Deployment Using Flask
How to deploy a flask application to Heroku via Heroku CLI and git
After we have created a user interface for loan approval prediction in the previous article, the app can be tested on our local computer to check for any errors and result in expectations. However, our goal has yet to be reached because it is not deployed. It means that the app must be running on the user’s computer. That’s not really efficient for business.
So, in this article, we will try to deploy our loan approval prediction in Heroku — platform as a service (PaaS). This article will cover how to register, the prerequisites, and setup preparation for deployment.
Buona lettura!
Our motivation
To optimize the operational activity in determining whether a customer who applies for a loan is granted or not, as Data Scientists, we understand that non-technical users don’t need to undertake technical activities. Thus, the application must be deployed on a server or cloud and non-technical users are given access to it. In the implementation, the loan approval verification becomes more efficient and operational cost is reduced significantly.
