EY brings more data into focus with Copilot for Microsoft 365

Many organizations find it challenging to track and make sense of all the data required to maintain and grow the business and meet tax requirements, so just imagine multiplying those challenges by tens of thousands. That’s the daunting task facing the EY Global Tax team. With thousands of projects and data stored in silos across complex organizations, EY relies on both technology and expertise to create informed insights to help its clients thrive in today’s rapidly changing environment. 

The first step in generating those insights is to bring the internal EY data from many sources and geographic areas into a single, unified view. Historically, EY Global Tax executives had to wait for internal data to be cleaned up, accessed centrally, and then pulled into a Microsoft Power BI report. That process could take a week or two, so by the time the data was ready, it was already at least two weeks old. If an error was found, the data had to be sent back to the originator to find the source of the error, correct it, and send the data back, creating more delay. 

To reduce turnaround time and frustration, EY continuously looks for ways to streamline its internal processes. In recent years, EY Global Tax teams began to focus on how new, generative AI–driven technologies could expedite data gathering and analysis for internal processes. The organization expects to use those learnings to solve for similar data needs for its clients.

“We’ve spent the last several years, possibly almost a decade, reorganizing our operating model to focus on being prepared for technologies like Copilot generative AI,” says Darren Beardsley, EY Americas Tax AI Leader. “And we’ve done things like standardize our business processes. We’ve centralized, we’ve organized our data, and we focused a lot on data governance.… So we're very excited that now the technology is here to be able to allow us to harness the power of the data that we've been focused on.”

Adding a smart digital companion

In late 2022, the EY Global Tax team learned about the soon-to-come Microsoft Copilot for Microsoft 365. Copilot provides an AI-powered productivity layer that users access from within Microsoft Teams. They can run a natural language query across multiple internal data sources across the enterprise to create a single, unified view of results. Copilot can summarize data, generate first drafts of reports, and unlock the full potential of data to accelerate the path to new insights. It not only reduces manual search time, but also harnesses a powerful language model to provide information in a more usable context. 

“The first time I heard about Copilot was a year ago when Microsoft announced it to all their clients, and my mind was blown away with the power that it can bring and the changes that it's going to make in our tax organization,” says Coskun Cavusoglu, EY Americas Financial Services Tax AI Go-to-Market Leader.

As part of its collaborative relationship with Microsoft, the EY Global Tax team was invited by the Copilot product team to preview the new software. The goal was to see how the team could apply a generative AI solution across the Tax organization. 

“Generative AI is all about unleashing our human potential, enhancing our capabilities, and highlighting our technical expertise, and we're going to get to do that all inside of Teams,” says Marna Ricker, EY Global Vice Chair—Tax. 

Redefining the work experience

The EY Global Tax team collaborated with Microsoft to build a proof of concept for a plugin that is used to directly ask questions through Copilot to expedite analyses and report creation. The plugin—tested on a developer tenant provided by Microsoft—redefines the report creation experience for EY Global Tax senior executives. Instead of using Power BI and waiting for data to be populated, they can ask questions directly through Copilot to get the information immediately. The proof of concept demonstrated the ability to seamlessly extend Copilot using Microsoft Graph connectors and EY proprietary sample data to bridge connections and insights across various business disciplines, using the security technologies built in to the Microsoft platform. 

EY executives can now have the power of real-time business insights, through simple, natural-language queries in Copilot. “Having access to all that data through Copilot and Teams is now just a matter of seconds for them to be able to just go in, open up Teams, and as they're chatting with maybe a senior executive, they're actually chatting with Copilot, which knows our language, speaks our way, has our DNA embedded in it, and it's basically directly coming back at them with answers that they're looking for,” says Cavusoglu.

Just by prompting Copilot, EY Global Tax executives can quickly assess their internal data to determine where to focus attention: on billing and collections, the health of engagement economics, the utilization of people according to their skill sets, hiring decisions, or the next big initiative. And then, depending on the results, users can generate tasks such as creating report drafts, schedules, follow-up meetings, and email communications. 

“I was very surprised with what's available out of the box. It does a lot of things that you need by just connecting it,” says Cavusoglu. “So what we had to do is we had to create an API that essentially brought the data together so that Copilot could talk to it. And once we did that, one of the beauties of this technology is that I didn't have to explain a lot of the terminology.” 

While testing the proof of concept on the Microsoft development tenant for various internal scenarios, the EY team also provided feedback to Microsoft about what works, what doesn’t, and how it could be improved. Once EY Global Tax has proven the value of Copilot for Microsoft 365 within its own organization, the team will start introducing its clients to how they can use it for their own internal data reporting.  

“As we've been working with Microsoft on this, we really have a good game plan and we know what not to do,” says Cavusoglu. “We know how people need to start thinking about it, how to get their data into the right place, and how do they serve their data securely.”

Based on those insights, EY is starting to strategize with clients about how to make sure their people and their data are ready to make the most of Copilot, and what clients need to think about if they want to build their own line-of-business applications around it.

Addressing security concerns

Security and privacy were, of course, crucial considerations for EY Global Tax in deciding how and where to use Copilot. Enterprise users want to know who has access to the information collected through Copilot and where that data is being sent. For the proof of concept, security built into Teams and Copilot followed all the policies established by Microsoft 365. The sample enterprise data used in the proof of concept to generate informed responses is present only as part of a prompt to the language model. The prompts aren’t retained by the language model and aren’t used to train it. Because all retrieved information is provided to the user based on their data access and permissions while signed in to Copilot, users can access only the data they’re authorized to use for their job.

Unleashing the power of more data

In the past, the sheer volume of data involved in client engagements resulted in a long lag time between accessing data and developing trends. EY Global Tax is looking to prove whether using Copilot together with Teams will help EY professionals build broader insights much faster.

“We've always had a challenge tackling large data sets in a rapid fashion,” says Shawn Smith, EY Americas Financial Services Tax Leader. “The ability to access data very quickly and get quicker insights for our teams and for running the business will really help us be more nimble and effective in the marketplace.”

Revolutionizing the way of work

EY Global Tax and many of its clients already use Teams, so they expect that having Copilot at their fingertips at all times will create new ways to collaborate and communicate with clients and coworkers. That will ultimately make meetings more productive. 

“Using Copilot in Teams is going to revolutionize the way we work because we now have this digital resource sitting next to us at all times of the day working with us to find the right answer,” says Cavusoglu. 

If an EY professional is meeting with their team or with a client and a question comes up that they don’t have an immediate answer for, they can query Copilot and get immediate input to formulate an answer. The conversation can continue, rather than having to say, “Let me get back to you on that,” which will help them meet their client's needs faster. 

EY Global Tax executives expect generative AI to transform the tax industry through automating many routine tasks and sparking new ways to look at data. “Generative AI is really transforming tax. It is automating, it is prompting, it is imagining, it is creating,” says Ricker. “The thing I probably love the most about it is that it is unlocking data in really unimaginable ways.”

Note: EY built this proof of concept on a developer tenant provided by Microsoft.