Microsoft Fabric – Experiences with implementation and utilisation

For the past couple of years, Microsoft’s flagship solution for data-driven decision-making has been the Fabric product suite. We have summarised the strengths and limitations of this rapidly evolving platform, which are essential to consider when selecting a data platform.

Growing confidence in an evolving solution

Microsoft Fabric combines new components and familiar tools such as Power BI. Since the first implementations, the platform has developed rapidly, significantly changing experiences and opinions. Initial curiosity has now turned into confidence, with many of our experts already swearing by Fabric, especially for simpler use cases. However, complex environments may require more creative and pragmatic solutions until Fabric’s fast-developing features reach an “enterprise-grade” level.

If your organisation is considering adopting the Fabric platform or planning to expand its use, the following points are worth noting:

Ease of use and compatibility

• Data management becomes easier when the entire data pipeline is within the same product.
• Fabric integrates well with Microsoft’s product family, particularly with D365 CRM, ERP solutions, and SQL Server, especially when using Fabric Link.
• Integrating source systems that provide REST APIs is straightforward and quick.
• Notebook tools are reusable and user-friendly, improving efficiency.
• The environment is already familiar to Power BI developers.
• DBT works reasonably well with the warehouse side.
• An AI assistant is integrated into the platform and is continuously improving.

Challenges and limitations

• Fabric’s security features are not yet fully developed, making it challenging to meet high-security requirements.
• Implementing Fabric in complex and large-scale environments with multiple databases or intricate data pipelines can be labour-intensive. This may lead to projects requiring extensive customisation and being difficult to maintain.
• Version control and change management in Fabric are still lacking, complicating the management of large projects. For example, using Azure DevOps is not straightforward.
• Connecting on-premise source systems (e.g., Oracle databases) requires significant effort, and maintainability is not yet seamless.
• Fabric’s pricing model can quickly lead to high costs, especially when processing operational data. However, relatively large environments can run on minimal capacity by optimizing solutions. Moving Power BI to a shared environment ensures that the limited Fabric capacity is sufficient for the data side.

A good choice, especially for Microsoft-centric companies

Fabric’s development roadmap aims for large, comprehensive solutions and is expected to mature rapidly soon. However, the platform is still evolving, and many features are yet to come. This requires patience in project execution and monitoring new product releases and updates. On the other hand, Microsoft’s significant investment in Fabric increases confidence in its long lifecycle, making investment decisions easier.

At Digia, we can also recommend adopting the Fabric platform more broadly for most of our clients – especially for smaller projects where data utilisation is limited to a few functions, and for companies that already use Microsoft products extensively.

For larger, more complex projects with higher security requirements, we still recommend more mature solutions such as Snowflake or Databricks, which have been on the market longer. However, Fabric is rapidly maturing to compete in this category as well.

If the development of a data platform and data-driven management are timely for your company, watch our Fabric webinar recording (in Finnish), or contact our experts.