Business Intelligence Developer
COOP's data warehouse environment demanded that new requirements be fitted into an existing setup where the data foundation and reporting had to align stably in practice. I worked on-site in Greater Copenhagen (Albertslund) as a Business Intelligence Developer in an environment where Teradata was the data platform and MicroStrategy was a central part of the BI layer, and where deliveries were largely driven by cooperation with the data warehouse team and the development department for the new checkout system (receipt system), ensuring modeling and execution were solved as something that could actually run and be used, rather than a theoretical model on paper.
COOP's BI setup needed a stable data foundation for analysis and reporting, where new needs had to be accommodated within an existing data warehouse environment without the solution ending up as a fragile layer of special logic that only worked as long as upstream systems and data deliveries never changed. The goal was therefore to make modeling, transformation logic, and consumption align so that deliveries could be reused and further developed over time.
The work took place within COOP's data warehouse context with Teradata as the platform, ANSI SQL for data processing, and MicroStrategy as a central part of the reporting and analysis layer, which in practice meant delivering models and a data foundation in a format that could be both executed stably and used consistently in reporting, without building solutions dependent on manual processes or individual knowledge.
A significant part of the task was linked to the receipt system (bonsystemet), where new development was delivered and where responsibility covered both data modeling and execution, ensuring the solution was both well-thought-out in structure and could realistically be executed in practice as an operational part of the BI delivery, instead of being reduced to a loose coupling that required subsequent firefighting to function stably.
As part of the work, professional support was provided for building a forecasting model that used double logarithmic exposure of historical data to predict how much would be sold of various products based on past sales, providing a more methodical approach to estimation without the need to invent figures for model quality or precision.
The delivery was solved on-site in Greater Copenhagen (Albertslund), with primary coordination through COOP's data warehouse team and the development department working on the new checkout system (receipt system), while contact with the business was typically secondary and went through the technical teams, which suited a task where the data foundation, modeling, and execution were the core elements.
Coop Danmark A/S is one of Denmark's major grocery players and operates Kvickly, SuperBrugsen, Brugsen, and 365discount, among others. Ownership lies within the consumer cooperative, historically known as FDB (The Danish Consumers' Co-operative Society), which is today organized as the association Coop amba, where the members are owners, and where Coop amba together with OK a.m.b.a. owns Coop Danmark A/S.
Official site: https://coop.dk/