Data science processes only provide usable information if the basis is right. Inaccurate data always means inaccurate insights. Or, in other words, if the walls on the ground floor are crooked, the building plans for the roof extension are useless. Anyone who continues to build makes it all the worse because in the end they can no longer keep up with the concealment.
Energieversorgung Mittelrhein (EVM) developed a desire to be able to understand the customer journey of customers in a more targeted manner. Based on this, the program included an automated and customer-specific approach based on contract and behavioral data. At the end of this chain of data-driven insights, forecasts and trends should finally be made analysable and used as economic decision-making tools.
How does data become a 360° customer view?
Which technologies support data science initiatives?
How can detailed forecasts be made?
In two consecutive workshops, evm and taod analysed the current situation and identified initial use cases. This resulted in an MVP approach, which is further developed in an agile sprint model based on the existing data strategy.
A record linkage model was used due to inadequate data quality in the billing system. It recognizes similarities between data sets using statistical methods such as Expectation Maximization — supported by a scalable cloud infrastructure.
Azure Synapse enables evm to perform powerful data processing through ETL pipelines and Spark clusters. Customer identification was automated, results were transferred to the Modern Data Warehouse and integrated into operational monitoring.
With taod, evm successfully cleaned up its complex database using record linkage. The new Azure-based Modern Data Warehouse provides the scalable basis for all of the company's future data science initiatives.
Together with taod, Energieversorgung Mittelrhein has succeeded in the complex and time-consuming adjustment of its database. Using the record linkage process, the cleaned data was then compared in order to build up a corrected history. The development of a new cloud-based modern data warehouse in Azure also provides a reliable and high-performance basis for all future data science initiatives.
Energieversorgung Mittelrhein AG wanted to better understand the customer journey of its customers and create a reliable database for this. The aim was to make contract and behavioral data usable for customer-specific communication and to make forecasts and trends analysable as economic decision-making tools.
Prior to the project, evm's database was too imprecise and fragmented to use data science methods reliably. Inadequate data quality in the billing system and several isolated data sets per household, which made a consistent customer view difficult, were particularly problematic.
Together with evm, taod developed an MVP approach for data science initiatives and used a record linkage model for data cleansing and customer identification. This was supplemented by building a cloud-based modern data warehouse in Azure Synapse with ETL pipelines, Spark clusters and Microsoft Power BI as part of the technological basis.
The introduction began with two consecutive data thinking and data discovery workshops, in which the current situation and initial use cases were analyzed. The record linkage process was then implemented, the Azure Synapse architecture was built and the adjusted results were transferred to operational monitoring and to the new modern data warehouse.
By matching isolated data sets, evm obtained a 360° customer view as the basis for personalized communication and valid forecasts. Even with millions of data points per process, the new architecture processes analyses with a latency of less than one second, and the record linkage model automatically cleans up more than 100,000 customer records. In addition, the new Azure-based cloud stack creates the basis for further use cases such as churn prediction, forecasting models and dynamic scoring methods.
Let us answer your questions during a non-binding initial consultation.
Make forecasts, automate your processes and uncover trends and patterns with our AI consulting.
With Microsoft Azure Services and our Azure Consulting, you can expand the limits of your current platform.
How to identify and promote the potential of data initiatives and accelerate them for data science solutions.
