 Welcome back to SuperCloud Six. I'm Paul Gillan here on theCUBE. And at SuperCloud Six, we've been talking about AI innovation. And one thing you've heard through all the interviews, I'm sure, is the importance of data, having your data act together, good governance, good clean quality data, not as simple as it sounds. One company that has been synonymous in the financial services industry with high quality data is Capital One. So a company that had modest beginning started out as a credit card provider and has grown to become one of the largest financial services firms in the world. And a lot of that is because of its innovation in data products, its ability to personalize and target its products to specific people and professions and industries. And one of the reasons it does that is it has very good governance principles. Christina Agia joins us. She is the enterprise, excuse me, Vice President of Enterprise Data at Capital One. And Christina, it's a pleasure to have you with us. Yeah, I'm thrilled to be here. Thanks for having me. So we've all heard of the term Chief Data Officer. You are Vice President of Enterprise Data. What does that role entail? Awesome question, and I love to start this way. So as a Vice President in Enterprise Data, my team really gets to focus on all of the experiences that support the data consumers of the company. So the analysts, the data scientists, the engineers, the product managers that are using data to power business decisions, they leverage a portfolio of products and platforms that my team gets to build and improve that help them on every step of the journey from finding data to actually understanding it to then using it. And that's everything from our marketplace for all of our available data to platforms like or Lake to our actual consumption tools. Now, one concept I know you have there at Capital One is that of data as a product. This is a concept that emerged from the data mesh concept that you've embraced. What does that mean, data as a product? Yeah, that is one of my favorite things to talk about and data products are so much at the heart of what our team gets to work on for Capital One. So for us, as you said, data products at their core are the philosophy of treating data itself as a product, applying product management principles to data. And what that means at Capital One is that we focus, as you would with any great product on the end customer that's using the data and building and curating data assets, whether those are APIs, real-time data streams, analytical files and tables and curating these reusable high-quality data assets that are ready to use for users across the organization. And the key thing here is that we work backwards from a customer need and package that data up in a way that it will be useful and reusable to break down data silos and just make it way easier to find and understand and ultimately develop our data-driven experiences. A company like yours is dealing with enormous volumes of data from all of your transactions, your customer accounts, just overwhelming volumes of data. How do you get a control of all that, get your arms around it and get it into one usable form? Yeah, gosh. And this is really a place where the cloud changed everything, right? As we've moved to the cloud, we have not only way more data, but way bigger varieties in data. We have increasing volumes of unstructured data. We have a tons of proliferations of formats and we get that data way faster. It's increasingly real-time. And so for us, the foundation of being well-managed is where everything starts. We have to know what data we have and where it comes from, why we have it and who can use it. And as we have scaled our volume of data, we've consistently true back to really managing data at every step of the life cycle and building scaled products and platforms that help the company detect, understand and protect data at every step in its journey. One sort of philosophical discussion we hear often is the merits of centralized versus decentralized data. How centralized are your data stores? It's a great question. So I would say that our philosophy around centralization and decentralization starts with the notion that you have to define central standards and requirements. We have to have sort of the same rules across the company for what needs to be true for that data to be well-managed, discoverable and accessible, but we absolutely subscribe to the notion of decentralized ownership, right? So the idea that across the business, we've got tons of different business owners that are best positioned to own and manage and package their data, build data products that we ready for use. So for us, we have central teams like mine that get to build the products and platforms that support core data management functions like data discovery and cataloging and data storage and access. And we work with teams across the company to leverage those products and platforms to make their data available. And much of the focus in that is one of really standardizing and building platforms that can help us automate and scale the common management requirements for data while letting business teams across the company take ownership and accountability for the actual data itself, curate it in a way that will work for their customers and be accountable for who is using it and for what reason. And you talked about standards and I remember from an earlier conversation we had, you talked about how you standardized the date format across all of Capital One, which it seems like a very small, minute detail, but it's very important to getting consistency. What kind of practices do you recommend organizations apply to maintain data consistency and standardization? Yeah, it's a great question. And on your point on dates, I lost probably weeks of my life to ISO date formats and the many different systems and formats and naming conventions that exist, especially when you're dealing with, as we've talked about right, a company that shifted to the cloud but had previously a lot of mainframe data systems as probably a lot of folks watching this do when you have this proliferation of data sources, even simple things can quickly get really complicated. And so for us, when we think about standards, we're talking about standards really in three different dimensions. One is the date example, standards around our data itself. And so we really strive to establish standards that promote consistency and usability and interoperability for our most ubiquitous data. So we have codified a set of requirements for data itself for that most common, most basic data that everyone across the company can and should be using. The second category of standards are really the requirements around our data processes and setting standards in this case, like data management practices and processes for what needs to be true for data at every step. So standards around how do you record and catalog the data we have, how to capture lineage, how to capture data quality rules and building products and platforms that embed those standards, whether it's the data definitions or the data management processes into the experiences that users at the company are leveraging to find, curate, produce and access data. And all of that is really on a foundation of some actual data management standards themselves. So things like standardizing the underlying formats of data in our lake or driving consistency in the sort of like interface designs and schemas that we're building. So we think about standards really across those three different dimensions. And then we really strive to build products and platforms that are gonna bring together and embed standardized definitions, standardized processes and standardized data management requirements to really make data consistent and usable across the company. Very disciplined approach. I imagine there's a lot of automation involved. Where do you apply automation to these processes? Yes. Automation is very much like the mission and vision for all of our teams. And it really starts with standards because you can't automate what you have in standardized. It's really difficult to automate a bespoke, unique, non-inconsistent process. And so standards are where our automation aspirations begin and we work to embed that automation as far left in the data lifecycle as we can. So I think what you have seen and what is true at other companies as well as we've done some research that you and I previously talked about is that companies are increasingly shifting as far left as they can in the data lifecycle to embed that automation and standards enforcement. And so for us, a lot of that starts at the moment data gets created, automating the processes to inventory and register it, to publish it and ensure it's of quality and all that goodness. Data, speaking of data quality, how do you ensure that? Who's responsible for data quality? Is there an individual or a group or a department? Yeah, it's a great question. Data quality, especially at a bank is so much at the core of our business processes. If we don't have quality data, we can't run many aspects of the business. So it's very much the responsibility of every leader in the company to take accountability for their data and its constraints. Now, where our team plays a really big role is in building both those governance frameworks so that we define standards that are consistent across the company data quality rules and also build tools that help embed data quality into all of the experiences that our associates are using. So it's very much a shared responsibility, but our enterprise teams play a key role in building the products and platforms that actually embed those data quality requirements into the ecosystem. And at times, for really important data, we might even step in and set sort of a ubiquitous data quality standard for the firm for some of our most important, most critical data. You spoke earlier about a marketplace. And I know one of the philosophies of Capital One is that data should be accessible. It should be self-service. How do you enable self-service access to data while keeping in mind you're in a regulated industry? There are a lot of constraints on what people can do with data. How are those guardrails applied and how do you make it easy for the people who need data to get to it? Yes, that is a great question. And there is this really important balance that you're teeing up between how we offer sort of seamless and self-service access to data while balancing security and being well managed and this sort of inherent tension that exists between the two. For us, the answer starts with a central self-service platform where any associate across the company can discover all of our available data. And we offer this central catalog is one way to think about it that lets users discover all the data that's available, get recommendations around the data that's most relevant to them. It's the place where we promote those data products that we talked about before and put information in the hands of every data user at the company so that they can find the data most relevant to their use case and request access to it. Now, we do operate this marketplace in partnership with all of the data owners of the company and so we offer them experiences to then manage access to their data and kick off experiences to send users into an actual kind of consumption experience to then integrate and use it when the right access is in place. It's sometimes said that culture eats technology for lunch. One of the competitive advantages I believe you bring to the market is you have established a culture at Capital One where data is a prized asset. For organizations that are still struggling with that, do you have any advice on how they can build a data-centric culture? I totally do and your one spot on that Capital One is very much a company where data is at the core of our founding. It is the founding principle of the company is using data to drive decisions and so it's very much embedded in how we operate and for us and others who want to operate in this way, my best advice starts with being a demanding customer of data as a business leader or product manager or engineer yourself. We should all be using data to guide our decisions and engaging deeply in the subject matter of the data that powers the products and experiences and business decisions that we all get to make. So step one is really like as a leader, be a demanding customer of data for yourself and your team, that's the first thing. And then you have to invest in that culture not just from a leadership level but with training and tools and support. Gosh, the world around us is changing so quickly. I'm sure that's a common theme in a lot of the conversations you all are having here where you have to invest in training and resources so that folks can constantly be upskilling and understanding where the latest innovations are heading and enforcing the importance of that with your teams and associates. And so when you build as a company this combination of real reverence for data in every leader and the supporting ecosystem of trainings and tools for us, we have self-service kind of data academies that folks can attend. Those things come together to really build a strong data culture for your organization. Now, SuperCloud 6 is all about AI innovation of course and you've got a strong data platform to work from. How is Capital One leveraging this data platform as it moves further into AI? Absolutely, it is so exciting, right? As you said at the very beginning at the heart of our AI aspirations is the need for excellent and well-managed data. And so the AI revolution is just raising the stakes on how companies manage and use data and amplifying the need to be well-managed, the need to deeply understand what data you have and where it is and what the source of truth for your data is that you can rely on. And so for us and the rest of the world as we're preparing for this new frontier we are getting ready to unlock all of the value of this increased volume of data we have. And we're doing that principally with really strong data management that is embedded and automated in our products and platforms. And that as we just talked about really great data culture where associates across the company are investing in understanding data, the innovations in data and how to unlock its full potential. So that's like there's a beautiful kind of virtuous flywheel that kind of comes together here of awesome data, powering excellent AI generating even more data, right? That then feeds back into that loop only made possible on a foundation of really strong well-managed data principles and platforms. You said everyone's excited about AI as you look out over say the next couple of years where do you see the big opportunities for AI influencing Capital One's business? Yeah, so as we said previously, gosh, data and data driven decision-making has been at the heart of how Capital One has run its business for a long time. And so as a result, AI and ML, they're really foundational in how we run our business and improve our experiences. And so from that perspective, AI development is no different whether we're trying to help customers shop more safely or giving them more timely insights into their finances. All these advancements in AI and the underlying investments in our data agenda just help us unlock new opportunities to deliver more personalized, more intelligent, more real-time experiences. And that's really at the heart of our vision to just keep making banking easier and better for all of our customers. So AI and ML are almost at the core of how we've always operated in that they are this next iteration of levers to get the most possible value and unlock out of our data to deliver the best possible experiences to all of our customers. Well, I can tell you, I mean, Capital One's slogan is what's in your wallet. And I can tell you that what's in my wallet right now is not one but two Capital One credit cards. And it is the best online customer experience I have ever seen from a financial services company. You really are delivering on your promise. Oh, we so love to hear that. And you can count on it just continuing to get better. Christina Egea, Vice President of Enterprise Data at Capital One, so happy to have you here today. Fascinating discussion and continued good luck. We'll be watching closely what you do in the future. Awesome, yes, thank you so much. And thanks again for including us here. I'm Paul Gellin, this is SuperCloud Six on theCUBE. We'll be right back.