Microsoft Power BI Copilot - Future is here

In March 2023, Microsoft announced the implementation of ChatGPT AI into their products, naming this functionality Copilot. It is a logical result of cooperation with OpenAI company, which created ChatGPT AI. Part of this announcement was a list of applications where Copilot will be available, and Power Platform was included. However, Power BI was not on this list, and from Power Platform family only Power Apps and Power Virtual Agents were mentioned. This was quite surprising because the Power BI service is already able to provide data insights based on prompts in natural language, so improving this functionality by ChatGPT capability seems to be a logical step.

Anyway, now, two months later, Copilot for Power BI will be announced and will be one of the hot topics of Microsoft Build event that will happen from May 23 to May 25.

You will find the Power BI Copilot announcement video at the end of this article, but first, let us summarize what we can expect from Copilot in Power BI.

Data to insights

The main goal of all data analysts is to help business users move as quickly as possible from data to insights. Data analyst can prepare beautiful and complex reports, but it is impossible to prepare our reports for all scenarios that could happen and for all questions that business users need to answer. Here, Copilot, working with the full data model, can provide tailored answers and visualizations based on questions defined by the business user in natural language.

And what does this mean for data analysts?

Data analysts can now build reports according to the Pareto principle, which means that 20% of visualizations will cover 80% of business questions. The rest we can keep for AI that will build custom visualizations on request. I am already looking forward to my first answer to a business user… "Ok, this is an interesting question, but... you will probably not ask it very often, am I right? Could you ask Copilot to build this one-time analysis for you?" (Ok, I know, it will not be so easy because business users will need time to learn how to write questions the right way…but let me dream about it.)

Analysis on request

Finding key influencers will not be a problem for AI. Business users will see that a KPI is outperforming, this will be visible from the main report. But there could be so many influencers that impacted this specific KPI, and building all visualizations will be time-consuming and inefficient. AI will analyze data and trends, identify main influencers and provide visualizations proving this analysis.

Writing a summary of this analysis that can be copy-paste into an email and sent to management… this part is already a piece of cake for ChatGPT.

Data Analyst Assistant

Copilot can also assist with building data models. Yes, I know, we are experts, we do not need any assistance, and we can do it faster without AI. It’s true, but the main goal of Power BI was the democratization of data analysis or self-business intelligence, as it was called years ago when Tableau and Qlik started to invent this new approach that Microsoft with Power BI brought it to perfection. Copilot is not here to replace data analysts; it is here to help business users with basic analytical tasks without bothering data analysts, and data analysts can focus on advanced analysis and optimization of company data models.

This sounds very promising to me. Now let's see how Microsoft presents Copilot for Power BI in action:

Microsoft Power BI Copilot Announcement

Microsoft Power BI Copilot Demo

That’s it for today. I hope you enjoyed this article. Don’t forget to like it or share it with somebody who could be interested. :-)

For more interesting Power BI tips & tricks, tutorials and use cases you can subscribe:

LinkedIn Power BI Newsletter

Power BI Blog on Medium.com

Power BI YouTube Channel