Since Microsoft launch over five years ago, Power BI has been enthusiastically adopted by developers across the spectrum. Today they have over three million active developers on Power BI—who harness the power of data, provide actionable insights, and deliver cloud-native intelligent experiences. Today, over 15,000 independent software vendors (ISVs) and enterprise customers use Power BI’s embedded APIs to build analytic experiences into their applications. At Microsoft Build, they are delighted to celebrate the developers and shared some exciting updates for the next generation of developer experiences.
Announcing AI powered development in Power BI
Microsoft has announced two new capabilities in Microsoft Power BI that simplify how developers build their analytics through the power of AI. They had introduced a new capability that enables developers to save time and build more complex solutions using natural language to build DAX calculations instead of writing code. Furthermore, they have introduced new capability that leverages machine learning to analyze usage patterns to automatically build aggregations for optimal performance without the need for data engineers to optimize queries.
Natural language to DAX generation powered by GPT-3
Microsoft’s mission is to make data analytics accessible for everyone and provide a no-code and low-code experience for every aspect of data analysis. DAX is the expression language used by millions of developers today in Power BI, Analysis Services, and Excel to define calculations that can range from one line to hundreds of lines of code. By harnessing the Power of DAX, developers can create sophisticated calculations and business logic.
Microsoft has announced that they will be bringing the capability for customers to use natural language to describe what they are trying to accomplish and have Power BI automatically create a DAX expression for them. By using natural language to create DAX calculations, are further democratizing the ability to create sophisticated business logic without having to become a DAX expert. It will also help developers save time from searching, writing, and refining formulas. At Microsoft Build, it will demonstrate a sneak peek of our new no-code experience for generating DAX natural language that is currently in development.
Microsoft collaboration with OpenAI
To enable these capabilities, Power BI is leveraging OpenAI’s GPT-3 (Generative Pre-trained Transformer 3). GPT-3 is an advanced natural language AI model, trained with 175 billion parameters, that implements deep learning to be able to both understand and produce human-like text based on a prompt in natural language. Microsoft has a strategic collaboration with OpenAI to accelerate breakthroughs in AI – from jointly developing the first supercomputer on Azure to testing and commercializing new AI technologies.
Power BI is committed to Microsoft’s responsible AI principles that ensure the use of AI is fair, inclusive, reliable, and respects privacy and security. The use of GPT-3 within Power BI has undergone extensive training with built-in safety controls to ensure that no harmful outputs are generated. Furthermore, GPT-3 leverages user input to generate the best formula options that enables an AI augmented developer experience and Power BI developers to maintain complete control of which formulas are applied by selecting the expression from a list of generated options.
Announcing automatic aggregations in Power BI
Customers like Walmart are increasingly analyzing massive volumes of data in Power BI. Power BI introduced aggregations—which combine the best of Power BI’s in-memory capabilities with the big data capabilities of Azure Synapse Analytics or other big data sources to provide fast interactive analysis over trillions of rows of data.
Building aggregations in Power BI require a data engineer to analyze telemetry and design an optimized set of tables that enable the most common queries to be handled by Power BI’s blazingly fast in-memory Vertipaq engine and have the less frequent, detailed queries delegated to the big data store.
Microsoft has announced that the preview of automatic aggregations over any data source including SQL, Azure Synapse Analytics, and other third-party big data sources will be coming to Power BI in July this year.
This new capability uses a machine learning model to analyze the query patterns from Power BI to the big data store and automatically design and build the aggregations in Power BI that deliver optimal performance. The machine learning model runs in the background, continuously tuning the aggregations as usage grows and usage patterns change. Automatic aggregations in Power BI will enable all Power BI Premium customers to benefit from the performance acceleration that Power BI provides, without having to invest in expensive and time-consuming data engineering resources.
Announcing streaming dataflows in Power BI Premium
Customers want to work with data as it comes in, and not days, or weeks later. Microsoft’s vision is simple — the distinctions between batch, real-time, and streaming data today will disappear. Customers should be able to work with all data as soon as it is available.
Microsoft has announced that Power BI streaming dataflows will be coming to you this July in preview.
Streaming dataflows allows every business analyst to work with streaming data with beautiful, drag and drop, no-code experiences. You will be able to mix and match your streaming data with your batch data. Working with streaming data is no longer limited just to data engineers.
Users can connect to streaming data (Azure IoT Hub and Azure Event Hub by preview) and combine it with reference data to perform data preparation operations like joins and filters as well as time windowing aggregations (such as tumbling, hopping, and session windows) for group by operations. All of Power BI’s rich data visualization capabilities will work with streaming data just as it does with batch data today. Streaming dataflows is included as part of Power BI Premium.
Streaming dataflows in Power BI empowers your organization to:
- Make confident decisions in near real-time. Be more agile and take meaningful actions based on the most up to date insights.
- Democratize streaming data. Make data more accessible and easier to interpret with a no-code solution and reduce IT resources.
- Accelerate time to insight. End-to-end streaming analytics solution with integrated data storage and BI.
Customers like Grab have been leveraging streaming dataflows in private preview, testing, and benefiting from the power of streaming data capabilities beyond what current solutions can offer while providing business users with potential faster and better insights.
Announcing embedding in Jupyter notebooks
Microsoft has announced a brand new open-source Python package that enables you to tell compelling stories inside Jupyter notebooks. Now, you can provide live reporting, educational visualizations, or quick access to analysis and saved views in a production context. And the best part is you can start embedding in Jupyter notebooks now with a free trial of Power BI Pro.
By embedding Power BI into notebooks, developers and data scientists can use the excellent visualizations in Power BI to tell a clear efficient and accurate story over data.
Announcing deployment pipeline automation APIs in Power BI Premium
Power BI deployment pipelines make it easy for Power BI developers to enhance their productivity, deliver content updates faster, reduce manual work, and set up production-ready automation in minutes. Deployment pipelines are incredibly easy to use— through a beautifully designed visual experience that is intuitive to all Power BI developers regardless of their technical skill. One of the biggest asks we have had from professional developers is the ability to automate the process.
To facilitate this, Microsoft has announced that starting today deployment pipelines have added Automation APIs enabling developers to use tools such as Azure DevOps, GitHub, and Azure Pipelines to automate the deployment of Power BI assets as part of the existing app deployment framework.