Agenda:

• Understanding the importance of having a high-performant solution
• Where do I start? How to identify the poorly performing area(s)

• Why is data modeling so important?
• Star-schema vs flat table
• Costs of not having a star-schema in place
• Role-playing dimensions
• Bi-directional filters

• Understanding the core components of the VertiPaq engine
• Describing different data compression encoding types
• Understanding the concept of cardinality levels
• Describing Auto Date/Time and calculated columns caveats
• Demo: Reducing Data model size applying various techniques (splitting columns, changing data types, reducing the cardinality levels, etc.)

• Understanding the concept of aggregations in Power BI
• User-defined vs Automatic aggregations
• Demo: Create user-defined aggregations, make Power BI “aware” of it, and unleash the full power using the Composite model; Check if the aggregations are being used by Power BI

• Introduction to the data transformation process
• Best practices for transforming the data
• Using Power Query Editor for data transformation
• Understanding the concept of Query folding
• Describing the most common scenarios for breaking Query folding
• Visualizing data refresh process
• Demo: understand the importance of Query folding by showing the difference in performance between the queries that return exactly the same results: one with query folding, one without

• Introduction to DAX calculations as part of enriching semantic model
• Understanding how little nuances in the DAX query can produce a significant difference in the performance
• Describing most common mistakes when writing DAX calculations

• Understanding what affects the report rendering time
• Using Performance Analyzer feature to identify the slowest report elements
• Best practices for visuals rendering time
• Demo: redesign Power BI report page to perform 5x faster, without violating the original report design

• Understand the difference between Import and DirectQuery mode
• Understand use-case scenarios for DirectQuery mode
• Best practices for using DirectQuery mode
• Things to avoid when using DirectQuery mode

My name is Nikola Ilić and I make music from the data. I’m really passionate about everything related to data. I have multi-year experience working with various flavors of data, predominantly with Microsoft’s BI stack. Therefore, I will write in most cases about SQL Server, T-SQL, SSAS, SSIS, SSRS, Power BI, Azure, etc. But, since I’m really curious by nature and always trying to learn something new, I will also consider writing about other data-related tools and techniques.