EY helps clients create long-term value for all stakeholders. Enabled by data and technology, our services and solutions provide trust through assurance and help clients transform, grow and operate.
At EY, our purpose is building a better working world. The insights and services we provide help to create long-term value for clients, people and society, and to build trust in the capital markets.
Professionals from EY and Microsoft explore the importance of data governance in the era of artificial intelligence (AI), highlighting the need for data integrity and trust to ensure reliable AI outputs. The conversation underscores the balance between data defense (risk management) and data offense (business enablement) to support innovation without hindering it.
Speakers:
Ryan Duffy, EY FSO AI & Data Strategy Leader; Senior Manager, Technology Consulting, Ernst & Young LLP
Effie Kilmer, Director, Purview Data Governance, Microsoft
Key takeaways:
The importance of data governance: In the era of AI, there is a high need for data integrity and trust.
How people and culture determine data success: Successful data governance requires a strong plan. Technology alone is not enough; organizations need to bring people along on the journey and upskill them to use data governance tools effectively.
The role and balance of data: There is a necessary balance between data defense and data offense. Speakers highlight the importance of having a solid data governance framework to support innovation without hindering it.
For your convenience, full text transcript of this podcast is available below.
Kathy Hevland
Did you know that 48% of workers worry the data they are using may be unreliable? Hello, I'm Kathy Hevland from the EY Microsoft Alliance. This is Tech Directions, where EY and Microsoft professionals explore transformative cloud solutions. Today's conversation is all about the data. And how, in the era of AI, data integrity and trust is more important than ever. Let's hear what leaders from EY and Microsoft have to share on the topic.
Ryan Duffy
Hey, Effie, I appreciate you being here today. I'm excited to talk about my favorite topic, data governance, data management, particularly for the financial services industry. You want to tell me a little bit about your journey, and how you started in with Purview?
Effie Kilmer
Yeah, thanks, Ryan, and thanks so much for having me. It's good to talk to you about data governance, especially because it's the most important topic of conversation now with the era of AI.
Duffy
Absolutely.
Kilmer
But I, randomly, started this journey about 10 years ago, working on a Power BI report, and really appreciating data quality, and the importance of traceability with data. And then, now, landing in my current role, I lead our product growth and partner enablement for Microsoft Purview Data Governance. So it's been a really exciting journey, and evolution of data governance, from 10 years ago to now, seeing the risk management function. And, now, it's really about business enablement and unlocking business value.
Duffy
Yeah. The defense versus the offense side.
Kilmer
Exactly.
Duffy
Yeah. They say, offense wins games, but defense wins championships. I think the same goes for data.
Kilmer
That's right.
Duffy
So in that evolution, it's not just technology. Processes have evolved. But I think people is a big component. Right? So what about data culture, or the core DNA of data governance, needs to happen to make it evolve in an organization?
Kilmer
The people component here is essential. I think, if we look at the past, and why data governance programs have failed, effectively, is because they took more of a technology-led approach. And it's really about bringing the people along on the journey to help upskill them to understand how they use the tool. So the culture piece is critical to getting right, and really sets aside organizations that are successful with data governance versus not.
Duffy
You have the cool technology. And now, with AI, if you don't have a solid foundation, there's a lot of risk in even the AI outputs, right?
Kilmer
Yeah.
Duffy
So having that foundation in data governance, and embedding it in your day to day, really makes a difference.
Kilmer
Yeah, exactly. And you probably see it too, working with large clients, and especially in financial services, how culture has been such an essential piece for maturing the practice of data governance, and supporting now what we're seeing today with GenAI and AI. People are really understanding you have to have the data foundation in place.
Duffy
So, from your perspective, how is Microsoft trying to evolve data governance?
Kilmer
Yeah, we've completely re-imagined the Purview Data Governance experience over the past year. So the new experience is incredibly business friendly. So data consumers can come in, easily discover, understand, access, and trust the data that they're using for an AI program, or really any project that they have.
With data quality, it's like you no longer have to be a technical expert to apply data quality rules and policies onto your data. You can actually think about it in a much more business-friendly lens. So we have out of the box controls that you can apply as a data quality manager. And we don't just have it so that you can apply data quality at the physical data estate, which is what we've traditionally seen in the market. You can now apply it at more of the logical concepts, and then it gets to trickle down. Because in the world that we live in today, data is just constantly evolving.
Duffy
And expanding.
Kilmer
Yeah, expanding, and growing, and you have multi-cloud data estate. So how do you control all that data? It's not going to be possible to do it at the raw physical data estate level. So data quality, we're super excited about that feature and capability. And then, the federated data governance practice that we brought in Purview Data Governance, where you can bring your CDO into one tool, and help have your CDO activate their controls and compliance framework that they've created, based off of their requirements inside their organization. And then federate it off into the domain leaders and then activate their teams and the data steward. So what we're doing is, we're having one tool where your data consumers, your data stewards, your data owners, and your CDO, can come together, and just seamlessly activate your federated data governance practice.
Duffy
And one thing that we've done, between Microsoft and EY, is helped develop what we're coining data governance by design. And in my mind, perfect data governance certainly doesn't exist, but really good data governance is as intuitive as possible. So what you were describing, it becomes part of your BAU. It's not an additional activity in order to carry out some of these data governance responsibilities and capabilities. It's just part of your day. It's just you doing your job.
Kilmer
How it should be.
Duffy
That's right. But it takes time to get there. So how would you describe starting out on your journey?
Kilmer
Yeah, starting off on your journey, try to understand what's the most important data that you have in your organization. I think taking a tops-down business lens approach is essential. Because if you look at your entire data estate, it's going to be incredibly overwhelming to understand, where do I start? How do I prioritize on what's important for my team? And it's essential to think about, what are your business goals? Talk to your marketing lead.
Duffy
Great point, great point.
Kilmer
Talk to your marketing lead. Ask, "What are our goals for the upcoming year in marketing? What are our goals in the sales team or the HR team?" Think about, what are we trying to drive for the upcoming year? Then, what's the data that's supporting it? Start there, because that's clearly your most important data that you need to make sure that it's trustworthy, that it's accessible for the team, that it can be used to track on programs. And so, I would say, start with your business goals, and then attach data to it. Of course, in the financial services landscape, specifically, the regulatory demand is constantly evolving.
Duffy
Huge.
Kilmer
And huge. And why we see financial services being more the mature industry in the market when it comes to data governance, because they've thought it through over the past several decades. And so, I will say that starting with your business, and then curating your data there, and then meeting your regulatory requirements, start there. And then see how you continue to grow. Because you don't necessarily need to govern and curate your entire data estate.
Duffy
So maybe you start with what, like you're saying, those business objectives, trace it backwards.
Kilmer
Exactly. If we look at five years ago in the data governance space, one of the challenges was, how do we create a business case for data governance up to when we get budget approvals for the upcoming year? And I think, now, we're sort of over the hump there, with AI coming in the fold, and the appreciation for data governance and all that work. But it's the stitching together of your data governance practice, not just getting in the weeds here underneath, but understanding how your impact, as a data steward, or a data owner, helps support the greater business outcomes is critical.
Duffy
Yeah. So how do you find the balance between data defense and data offense, like we were talking about? Because it can't be binary. I think, traditionally, especially in the financial services, highly regulatory environment, highly compliance reliant activities. But, now, we're starting to see that shift to more of an offense. It's not one versus the other. So how do you find the right balance?
Kilmer
Yeah. In financial services specifically, I would say, data governance sat in that defensive position as a risk management function. And having the right framework in place is critical. And maybe your framework evolves. The controls that you have in place to make sure that you have ownership for all your data, that your data's properly classified, that you have glossary terms applied to them, data quality rules, traceability with lineage. So that's all in the defense. And then, balancing, building off, I would say, off the framework that you have for your controls and your compliance framework, then you can think about enabling your business. Because you have the confidence that you need to adhere to your regulatory requirements, as it continues to change over the years, but also not hindering innovation inside your organization. Because I think data governance has traditionally been no. So you hear data governance, and it's like, "Oh, we're slowing down."
Duffy
Thou shalt do this.
Kilmer
Yeah, we'll never get access to the data. And no one really understood doing this extra work would get me to X, help me with my daily job, or the business objective that I have. So there's a fine balance. But you have to have the framework and control in place. And then you can think about building an enablement for the innovation, more of the offensive side, which I think everyone's really excited about.
Duffy
Yeah, yeah, absolutely. So there was a recent EY survey that showed almost 90% of the respondents use some form of AI in their day-to-day jobs. But in the same survey, 48% worry about the quality of the data as an input, and how that's going to impact the trust in the data, the outputs, and really the benefit and success of the AI itself. So how is Purview helping to enable AI-ready data?
Kilmer
Yeah. Data quality's essential. So to have your data be reliable and trusted is what makes data valuable. So what Purview Data Governance does, it enables you to apply data quality policies. So not just applying policies at your data from the governance domain levels into your data products and your assets, but it also makes sure that the timeliness of your data is appropriate. But I would say that, from a data quality rule perspective, it definitely supports to make sure that your data is trustworthy.
And so, in a data product, you can see how the data is being used, data quality score, understand how that score is being powered. Because it's not just governing your data going into your AI models, it's also making sure that there's no bias out there on the other side. And so with Purview Data Governance, it's our ambition and target to help with data quality and reliability of data going into AI, but then, also, governance after.
Duffy
Absolutely. And I think one thing, key phrase you hear a lot, and people aren't exactly sure what it means, is human-in-the-loop, right? So a lot of what we're talking about, some of it is pipelines, running, data, controls, outputs. Where's the human, with the advent of AI, AI is not going to take your job. It's going to take your job description. The human hasn't gone away. We've become more efficient, and we're doing different things. How are you seeing organizations think about their data estates?
Kilmer
Data is the heartbeat of every organization. I think every large organization wants to understand, how do they squeeze their data to get value out of it and create it.
Duffy
Sure. Blood from a stone.
Kilmer
Right, exactly. And get the competitive advantage that they rightfully can have if they can understand their data better. But I think data is used in everyday programs, whether you think you're a data consumer or you don't. We say we're data nerds, but everyone's a data consumer. Organizations look at their data as a competitive advantage, how to build new products, build new revenue streams.
Organizations look at data as a means to better serve their customers, as a means to do better reporting up to their business executives to make more informed business decisions.
Duffy
Yeah. So, Effie, this conversation has been fantastic. I get fired up talking to you every single time. Like I said, data nerd to data nerd.
Kilmer
Me too.
Duffy
If there's one thing that you could take away from this conversation, and really let the audience stew on it, what would that be?
Kilmer
Two things.
Duffy
Oh, okay.
Kilmer
One, I would say, embrace change. Take on change as a means to educate yourself and learn what capabilities are out there in the market. So you can see what's the art of the possible out there. Because there's great innovation that's taking place in the data space, and in the data governance space specifically. But I loved our conversation before about, how do you get started with data governance? And I think it's super important for anyone getting started, just think about the people first and the process first. And then think about what the tech is. So practice first, tech enabled.
Duffy
Oh, I love that. That's great.
Kilmer
What would it be for you?
Duffy
I think the first would be, don't forget the fundamentals.
Kilmer
Yeah.
Duffy
Don't forget the baseline fundamentals. A lot of the data governance, and data quality, and data management principles that really spun out of 2008, like we talked about, particularly for the financial services industry, haven't changed. The data's changed. It's expanded. But, fundamentally, a lot of it has remained constant, and we're still evolving on it. So don't forget the fundamentals. And then I think the second is, I agree with you, embrace the change don't be scared of it. I see a lot of people starting to embrace AI. Data is everyone's job. No matter where you sit in the organization, no matter what organization you're in, financial services, consumer packaged goods, power and energy, the world is data driven now. And we all need to upskill ourselves, and not be scared of it, because that's going to be the future.
Kilmer
We all have a role to play.
Duffy
We have a role to play.
Kilmer
Great. Well, thank you so much, Ryan. Great chatting with you.
Duffy
Yeah, great chatting with you, as always.
Hevland
I hope you gained valuable insights for a successful data strategy from this discussion. For more insights from EY and Microsoft leaders, visit us on ey.com/techdirections.
In this video episode of the EY Microsoft Tech Directions podcast, the discussion highlights how technologies like Microsoft Copilot empower talented and diverse workforces.
In this video episode of the EY Microsoft Tech Directions podcast, the speakers discuss how to use AI and Microsoft Copilot to enhance sales processes and customer engagement.
In this video episode of the EY Microsoft Tech Directions podcast, the speakers discuss how to use AI and Microsoft Copilot to enhance sales processes and customer engagement.
In this episode of the EY Microsoft Tech Directions podcast, we discover how advanced analytics, AI, and a data-first strategy are transforming finance, accounting and assurance services at an enterprise scale.
In this episode of the EY Microsoft Tech Directions podcast, we look at how developing broad data strategies is crucial to transforming operations and driving innovation.
In this episode of the EY Microsoft Tech Directions podcast, we look at how best to navigate cross-border cyber security, focusing on digital identity and zero trust.
