Power BI and Generative AI: What You Need to Know

With generative AI tools taking the internet by storm, it was only a matter of time before SaaS tools started integrating the power of AI into their platforms.

Microsoft is no exception. They’ve added generative AI tools to the existing Power BI platform. These brand new features impact everyone who uses Power BI - whether you’re responsible for your database’s architecture or you’re an employee who regularly inputs data. Here’s what you need to know about the generative AI tools in Power BI and how they impact your data.

Copilot: Microsoft’s Answer to Tools Like Chat GPT

Copilot is Microsoft’s answer to tools like ChatGPT, but with a Power BI spin. It offers more tactical generative AI services. Copilot exists in two forms: standalone (which looks at all of your data and tools) and pane (which looks at only the current report you have open.) Some of the overarching features with Copilot give you the ability to:

• Build new reports quickly - and with high levels of specificity
• Generate visualizations from data
• Analyze data 
• Analyze existing reports
• Create narratives from data
• Develop forecasting

Let’s dive into the specifics of what these look like.

Copilot’s Capabilities

The chat feature within Copilot makes it easy to describe what you want, and the tool will output it instantly. This saves time when creating simple reports, data visualizations, and other reporting features.

• Ability to chat with your data: Ask for ad-hoc analysis or insights, create and analyze visuals, and get summaries of reports or whole topics. 

• Search for artifacts: Find items, content, and specific data points. Note: copilot is limited to only the information that you have access to, not necessarily everything in the database.

• Summarize a report or topic: Get a quick overview of an open report or synthesize multiple reports together into an overview.

• Sculpt reports: Create a new report, suggest content for a report, summarize the underlying semantic model, create a visual, write daxe queries and add descriptions for your semantic model measures.

AI Runs on Clean Data

The key to getting meaningful results from AI such as Copilot is having clean data. This is a two-phase process. The first phase involves collection the data from the various sources within (and perhaps outside) of your organization. This data then needs to be compiled and aligned into a meaningful whole. Along with this, data governance practices and safeguards need to put in place to ensure the data stays clean and accurate going forward.

Once this cleansed data is in place, a process know as grounding takes place to help ensure Copilot provides correct answers. Grounding may include:

• Limiting data items available to Copilot to ensure data privacy is maintained
• Creating synonyms for Copilot so it knows all the different ways the organization may refer to a given entity or calculation (profit and net income, for example)
• Providing verified answers to Copilot so it has examples of the correct way to derive an answer
• Create custom instructions for handling special situations.

Copilot’s Requirements

The major thing to know about copilot is that while it works within Power BI, you can only access Copilot with a Microsoft Fabric subscription. Fabric is both an add-on to Power BI and an extension of Power BI. We’ll explore Fabric a bit more in-depth below, but it’s important to know that you can’t access Copilot without Fabric. In addition, there are a few other requirements per Microsoft's guidelines to be aware of:

• Your administrator needs to enable Copilot in Microsoft Fabric.
• You need to have an F2 capacity or above to be able to use Copilot.
• Your Fabric capacity needs to be in one of the regions listed in Fabric region availability. If it isn't, you can't use Copilot.
• If your tenant or capacity is outside the United States or France, Copilot is disabled by default unless your Fabric tenant admin enables the tenant setting called Data sent to Azure OpenAI can be processed outside your tenant's geographic region, compliance boundary, or national cloud instance in the Fabric admin portal.
• Copilot in Microsoft Fabric isn't supported on trial stock-keeping units (SKUs) or trial capacities. Only paid SKUs are supported.
• Copilot isn't yet supported for sovereign clouds due to GPU availability

You can find more details in Microsoft’s support article: Introduction to Copilot.

Microsoft Fabric: All-in-One Database, Analytics, and Generative AI Powerhouse

In May 2023, Microsoft announced Microsoft Fabric - a new, all-in-one unified analytics platform that synthesizes data from Azure Data Factory, Azure Synapse Analytics, and Power BI and uses AI technology to automate processes and improve output. Fabric includes Copilot as an integrated tool, which is available in any of the core components:

• Data Engineering - Enables users to design, build, and maintain infrastructures and systems that enable their organizations to collect, store, process, and analyze large volumes of data.
• Data Factory - A modern data integration experience to ingest, prepare and transform data from a rich set of data sources (for example, databases, data warehouse, Lakehouse, real-time data, and more). 
• Data Science - Empowers users to complete end-to-end data science workflows for the purpose of data enrichment and business insights. You can complete a wide range of activities across the entire data science process, all the way from data exploration, preparation and cleansing to experimentation, modeling, model scoring and serving of predictive insights to BI reports.
• Data Warehouse - Provides customers with a unified product that addresses every aspect of their data estate by offering a complete, SaaS-ified Data, Analytics and AI platform, which is lake centric and open. 
• Real-Time Analytics - Reduces complexity and simplifies data integration. Quick access to data insights is achieved through automatic data streaming, automatic indexing and data partitioning of any data source or format, and by using the on-demand query generation and visualizations.
• Power BI - The world's leading Business Intelligence platform. It ensures that business owners can access all the data in Fabric quickly and intuitively to make better decisions with data.

Fabric also uses Microsoft’s Fabric with Azure OpenAI Service to allow users to create custom AI programs for their data. All of the above core components can utilize Azure OpenAI and Copilot AI.

Explore Fabric and the AI tools within in more detail in our article: Should You Switch to Microsoft Fabric? A Guide for IT Directors

Both Copilot and Fabric add in exciting new AI features to Power BI to make it even more powerful. 

For database engineers, architects and managers: you can create complex reports on the fly and troubleshoot points of friction faster than doing it manually. Think of Copilot like an analytics assistant. 

For users, you can answer questions and engage with your data quickly. Instead of needing to send a “what am I looking at?” or “how do I find…” email to IT, you can ask Copilot these questions directly and get an answer on the spot.

No matter how you engage with your data, these new generative AI tools within Power BI will save you time, effort, and energy.

Power BI and Enterprise-Scale Analytics with SCS

Superior Consulting Services (SCS) is a Microsoft-centric technology firm providing innovative solutions that enable our clients to solve business problems.

Contained within your data is the story of your organization. Effectively unlocking that story is the key to the success of any modern enterprise. We create a well-designed, appropriately scaled data analytics environment to reveal that story to personnel throughout the organization for faster, more informed decision making. 

We want to help you tap into the full potential of your data. Explore our custom training programs to get started.

References:
Microsoft Fabric Overview
Microsoft Fabric documentation
Overview of Copilot for Power BI

Superior Consulting Services (SCS) is a Microsoft-centric technology firm providing innovative solutions that enable our clients to solve business problems. We offer full-scale data modeling, analytics and custom app development.