By WalkingTree   August 05, 2020

Power BI

Various decisions have to be made while starting a BI project. The decisions include things like sourcing your data, where and how to create queries to cleanse, transform, and reshape the data. Most Power BI solutions progress through a few stages, which include:

  1. Creating a simple prototype to explore ideas and get feedback from users and business leaders
  2. Building a working proof-of-concept to frame the solution
  3. Designing a foundational data model
  4. Completing a production-scale solution
  5. Repeated enhancements & releases

Let’s take a look at some guidelines for designing power BI solutions. 

  • Right-sizing solutions – One of the first questions to consider is whether the solution you intend to create with Power BI is a quick informal project or a formal project. Informal projects don’t necessarily have to conform to all the design best practices. If your organization has a process of delivering and supporting company-sanctioned applications, then you will need a roadmap for transition from an informal project to a solution that the business is prepared to support.
  • Planning for separation – Separating the data model from the report files helps in focusing more on report design. The certified dataset is deployed, secured, and managed as a separate object. Reports are connected to a Power BI dataset and any number of reports can connect to a published dataset.
  • Source data – Data from source systems can be loaded into a managed database or data lake before using Power Query to load the data model. Data from files may be transformed, cleansed, and validated before it is ready to be loaded. Using “brute force” transformation steps to replace values and cleanse records one step at a time. 

 

Read on to know more about futureproofing Power BI solutions 

Blogs

Translate »