The purpose of this project is to deal with a real world clustering problem and approach it with a business concern: from explaining the insights we can get just from the data, find a ML model that would fit on the clustering problem and finally put into production the model and the insights we can get from it.
For this project I've chosen this Kaggle dataset. This dataset contains information of clients of a store: from personal information to the money spend in different sections of the store.
In the case of a business, how to take advantage of the information of your customers? Well, if we are able to segment our customers will also be able to develop a deeper understanding of them and develop specific marketing strategies for them.
Final results can be retrieved by POST. An API has been built in order to able request which group a customer belongs to based on their data in order they can receive automatically specific offers based on the group they belong. An example of a possible request and a response could be as the following:
Click here to properly visualize the internal links of the Jupiter notebook.