Olymel GPU

Unifying Olymel’s data to better understand its markets

Person using her olymel dashboard to better undertand markets

The context

Olymel is a Canadian leader in the production, processing, and distribution of pork and poultry products. The company operates an extensive network of plants across Canada and the United States and exports its products to more than 65 countries.

Its operations, supply chains, and customers generate significant volumes of data every day. Olymel wanted to make better use of this information to support consumer behaviour analysis and decision-making across the organization.

Project details

Industry

Agri-food

Technologies

Microsoft Fabric

T-SQL

Power BI

SODA

Azure DevOps

Services

Données et BI

Development

Design

The challenge

Olymel’s data was spread across multiple systems, formats, and data providers, including Nielsen, Costco, Loblaws, Sobeys, AS400 databases, and Excel, CSV, and Parquet files. 

Each source relied on its own product codes and business logic. This fragmentation made cross-functional analysis more complex and required teams to work with manually built reports, without a single source of truth or guaranteed consistency across divisions. 

The project also had to manage more than 70 tables, asynchronous refreshes, and frequent format changes while maintaining data quality and traceability.

The task at hand

We designed and implemented a modern data warehouse on Microsoft Fabric, structured around a four-layer medallion architecture: Bronze, Silver, Gold, and Platinum. 

Each layer progressively transforms raw data into enriched, business-ready tables. SCD1 and SCD2 mechanisms also preserve a complete history of changes. 

We then developed a product unification mechanism that connects Olymel’s internal product codes with universal UPC codes. The SODA library was integrated into the pipelines to automate quality controls and regression testing. 

Finally, we implemented a semantic layer and a Power BI Design System to help business users create their own reports more independently.

The end result

The data warehouse now allows more than 100 Olymel users to explore data and build their own analyses. 

Teams have access to a single source of truth for operational and commercial data. They can better analyze consumer behaviour and sales performance without relying on technical support for every new request. 

The platform also makes it easier to integrate new partners and data sources, giving Olymel a reliable and scalable foundation for future needs.