Predictive data models
In order to improve customer experience and reduce complaints in the automotive industry we've analysed unhappy customers. Various reporting/ data science tools have been used to present a clear picture to management on the way forward. The model is also used to contact potential unhappy clients before they lodge complaints.
A client required reporting tools to help the company interpret data more efficiently. We created a data warehouse with OLAP documents as well as reports on demand. We designed a star schema database and developed ETL scripts to migrate data into the data warehouse. Additionally, we developed custom reports in both PHP and KnowageBI.
Using Python and TensorFlow, we developed an image classification system. In future, this system could be used to score users on their photographic skills or to identify incorrectly classified images.