Speaker

Mathias Thierbach

Mathias Thierbach

Head of Finance Systems at YouGov Plc, Data Platform MVP, and Developer of pbi-tools

London, United Kingdom

In 2015, after having spent over ten years as a Software Developer and Architect with Microsoft technologies, Mathias Thierbach moved into the Microsoft BI space. He soon landed on Power BI, but also realized quickly that the development and engineering tools and practices were nothing like the ones well established in software development. This is how pbi-tools started as a project, the only complete source control solution for Power BI.
Today, leading a growing data management team at YouGov, he experiences the benefits of those efforts every day. Having open sourced the project in fall of 2021, Mathias spends a lot of his free time bringing source control and DevOps practices to the wider Power BI community now.
In addition to his open source engagements, Mathias cares deeply about his role as an enterprise technology leader. Like many, once having started as a single contributor technologist, he had to pivot significantly when he moved into a manager role, responsible for building, stabilizing and growing a team of data engineers, analysts and architects. Mathias is now passionately sharing the many experiences and learnings that came out of that journey with the community.

Area of Expertise

  • Information & Communications Technology

Topics

  • power bi
  • DevOps
  • Azure
  • source control
  • Mentoring
  • Career development
  • Automation
  • CI/CD
  • BI & Analytics

Incremental Refresh and partition management

You have lots of data, well Power BI can handle it. Build efficient loading data models by using Incremental Refresh in Power BI.

While setting up incremental refresh is easy, what is happening in Power BI.com is more of a mystery. This session does a deep dive into what happens when you are incrementally refreshing datasets.

This will be a technical dive in to Power BI Incremental refresh using SSMS, SQL Profiler, and Tabular Editor 3

Attendees will walk away with a complete understanding of the Incremental refresh set up within Power BI. Then a technical understanding of what happens at the Analysis services side inside PowerBI.com.

Extra topics also covered:
- Hybrid refresh policies
- Advanced features enabled via the Tabular Object Model
- Impact for deployment of datasets (DevOps concerns)

Power BI Report development in a team with pbi-tools

Power BI projects often start very small, with a single person quickly creating impressive results using Power BI Desktop and a project that is fully contained in a single file. When those early successes are then expected to scale in an enterprise context and to be replicated by a larger team, things get more complicated. There is no obvious scale-out path. And whilst for Power BI Premium and the XMLA endpoint model developers have some advanced third party tooling at their disposal, there is nothing comparable for report development. Report builders are limited to Power BI Desktop and black-box PBIX files.

This is where pbi-tools comes in to fill a significant gap. Mathias, the author of pbi-tools, will explain and demonstrate how report is lifted to a whole new level: with a source control compatible view of what's inside a PBIX file it is possible to review and check what's changed between versions, collaborate on the same project within a team, and even apply scripted transformations at the source code level.

Mathias started developing the tooling and associated workflows some five years ago and was able to test and improve the system over and over again since. Attend the session to benefit from that experience and take home lots of practical guidance and ideas.

The internals of a PBIX file and how pbi-tools helps make sense of it all

The PBIX file is the fundamental building block when it comes to Power BI development. You may have heard that PBIX files are simply ZIP files with an internal file structure, however, very little is publicly known beyond that.

Mathias, for his work on pbi-tools, has spent years deciphering the (undocumented) PBIX internals. Those efforts have led to pbi-tools, today the only end-to-end source control solution for Power BI development.

This session will explain what's inside a PBIX file and how everything is linked together. Using the folder structure view of a PBIX file provided by pbi-tools, all of the components (Model, Report, Queries, Resources, Settings) will be looked at in detail. You will understand how to track changes made to a PBIX file using Power BI Desktop, what those changes mean, and how to apply similar changes offline, targeting the pbi-tools folder structure only.

Using that understanding, the session will further look into more advanced automation scenarios which become viable once the PBIX internals are known and provide various examples. This session is likely to inspire you to take your Power BI projects to new levels of automation!

What is GIT and how do I get started? An introduction for Power BI developers

The session will explain GIT concepts to people who are new to source control or who want to refresh their knowledge. Power BI developers without an engineering background have traditionally felt a little lost in the world of GitHub and Azure Repos. However, tools like Tabular Editor and pbi-tools have brought source control capabilities into the Power BI ecosystem. There are even hints that Microsoft might be taking source control for Power BI more seriously now! Time for everyone to learn the basics, understand the key concepts and know which tools to use and where to find help!
If you want to understand what source control is and what terms like git, commit, branch, and pull request mean, this is the session for you. In addition to explaining the basics the session will also demonstrate some end-to-end development workflows using git for Power BI projects.

Advanced git Branching Model for Power BI

This session introduces in theory and practice a git branching model supporting collaborative incremental releases for Power BI datasets and reports.

The model is collaborative: it is designed for multiple team members working jointly on a shared set of changes.

It is designed for incremental releases: Those continuously deliver business value with a regular and reliable release cadence and allow the product manager, the develop team as well as stakeholders to refer to a particular feature set using a shared versioning scheme.

Using both the source control as well as deployment capabilities of pbi-tools (https://pbi.tools), this model was developed by Mathias as part of building and growing a distributed data management team and was continuously improved over multiple iterations.

Join the session to learn how it works, the thinking behind, and how to adopt it for your own workflows. Foundational understanding of source control and CI/CD is recommended.

Using pbi-tools for advanced Power BI CI/CD pipelines

pbi-tools (https://pbi.tools/), having brought end-to-end source control capabilities to the Power BI ecosystem, also allows the creation of advanced CI/CD pipelines for Power BI Reports, Dataflows, and Datasets.

Where you need to go beyond Power BI's own Deployment Pipelines or want to rely on automation with Power BI Pro, pbi-tools is your go-to solution.

This session gives you all the necessary bits and pieces you need to set up your own pbi-tools powered pipelines. It will cover deployment manifests, what type of source artifacts are supported, the concept of deployment environments, and how to integrate it all into team workflows. You will learn what it takes to set up CI/CD pipelines in both Azure DevOps as well as GitHub Actions.

Advanced deployment manifest features are also covered, for instance parameterization of deployment artifacts and environment properties. You will be able to take home a good understanding of all required ingredients, but most importantly a list of sample repositories and pipeline definitions to use as templates.

Power BI Model Development Best Practices in a Team

Power BI projects often start very small, with a single person quickly creating impressive results using Power BI Desktop and a project that is fully contained in a single file. When those early successes are then expected to scale in an enterprise context and to be replicated by a larger team, things get more complicated. There is no obvious scale-out path, particularly with regard to model development.
Mathias, who has been on this exact journey over the past six years, presents the practices he has developed during that time, going from a single contributor beginning to a highly effective team setup. The session covers processes, tools, roles and automation, with a focus on very practical take-aways.

The session presents and explains the components required to establish a highly effective team developing Power BI models: Processes, Tools, Deployment Automation, and Documentation. Attendees will be able to take away plenty of proven ideas and patterns, including a source control based development model, a change management system using release pipelines, and a CI/CD setup providing automation at all possible stages. The key focus will be on the right processes to implement within the team.

In addition to Power BI Desktop, tools covered in the session include Tabular Editor, DAX Studio, and pbi-tools.

Lessons learned building a Power BI Team

You have spent most of your career with a focus on technology and you're feeling at home in that domain. Now you have been asked to build a team underneath you to take your previous accomplishments to the next level. Or you are considering a career move and step into management. If that is you and you are a technologist who is now a new or aspiring manager, this session will give you plenty of practical advise what to expect and prepare for.

From his own experience of having gone down this path over more than five years, Mathias will present his learnings and insights. It turns out the field of managing people and ensuring a team can function well as a coherent unit is rather different from the field of technology and computers! This session will help you not to feel lost and overwhelmed and instead take on this new challenge with great confidence and joy. The main focus is on "soft skills", however, embedded in a professional Power BI context.

Most importantly, you will be able to learn from someone else's mistakes instead of repeating them all yourself! Or, you might instead decide that a move into management is not for you after all and you should rather stay in a single contributor role. Either way, you will walk away with lots of trips & tricks. You will learn techniques for having greater awareness of yourself and others. You will get a useful reading list. And you will get plenty of food for thought and reflection. Ideally, this will make some contribution to making you a better leader and manager!

Power BI Next Step 2022

September 2022 Billund, Denmark

Data Insight Summit

September 2022 Chicago, Illinois, United States

Mathias Thierbach

Head of Finance Systems at YouGov Plc, Data Platform MVP, and Developer of pbi-tools

London, United Kingdom