Capital One, ranked 82nd on the Fortune 500, has lengthy been acknowledged for its data-first method and early adoption of cloud and AI. Amy Lenander, Chief Information Officer, has spent over 20 years serving to form that evolution — remodeling the corporate from a data-savvy financial institution into a frontrunner in AI-driven innovation.
In the present day, Lenander leads Capital One’s enterprise knowledge technique, advancing its imaginative and prescient of utilizing knowledge to ship smarter, extra impactful monetary merchandise.
On this dialog with CDO Journal, she displays on the corporate’s knowledge transformation journey, the foundations for scaling AI responsibly, and the way a “data-as-a-product” mindset is fueling enterprise-wide innovation.
Edited Excerpts
Q: You’ve been with Capital One for over 22 years. Are you able to stroll us by means of the corporate’s digital transformation journey? How has your knowledge technique developed, particularly as AI turns into more and more central to the enterprise?
Information has all the time been on the coronary heart of Capital One. The founding concept of the corporate was to make use of knowledge and analytics to ship higher monetary services tailor-made to clients’ wants. Since then, the know-how to appreciate that imaginative and prescient has solely gotten higher.
We grew to become the primary U.S. financial institution to go all in on the cloud, enabling us to leverage knowledge and know-how extra nimbly and successfully than ever earlier than. This opened up a whole lot of alternatives for us in knowledge and helped us to develop a contemporary knowledge ecosystem that might be prepared for knowledge, machine studying, and AI at scale.
In the present day, with even larger advances in AI and much more knowledge, there are extra methods we will capitalize on our knowledge technique and ecosystem. It means we’re trying intently at issues like unstructured knowledge and making certain knowledge is out there in real-time so as to use AI fashions in real-time.
Tapping into our personal knowledge continues to be a bonus for us, and now we have now far more of it. Our method hasn’t modified, however execution has developed to maintain tempo with the rising quantity and significance of information.
Q: Having beforehand served as a CMO and a CEO of Capital One UK, how has your background on the enterprise aspect formed the way you lead in knowledge right now?
Throughout each position I’ve had with the corporate, I’ve used knowledge to drive enterprise choices. That have helped to construct my instinct on what sort of information is most necessary for enterprise groups throughout the corporate, which helps me perceive which capabilities are most necessary to concentrate on.
My expertise working within the enterprise additionally helps me have empathy for the urgent enterprise wants of our enterprise companions after I must ask them to prioritize one thing that’s necessary for us to ahead our enterprise knowledge technique.
Q: What impressed the launch of Capital One’s Chat Concierge AI for automotive consumers, and the way does it mirror your broader method to accountable innovation?
We constructed Chat Concierge as a proprietary multi-agentic conversational AI assistant for automotive consumers and sellers in early 2025. It could actually carry out duties like evaluating autos to assist automotive consumers determine on the only option for them and scheduling appointments with salespeople.
The framework we constructed to assist Chat Concierge will be prolonged to different customer- and internal-facing use circumstances. Every time we have a look at cutting-edge know-how, we should steadiness these alternatives to boost each buyer and enterprise experiences with a well-managed, risk-centered method.
Q: What are among the greatest challenges in deploying AI fashions at scale inside a monetary establishment like Capital One?
Even earlier than AI or generative AI, we’ve been leveraging machine studying at scale throughout our enterprise. We all know that in the case of addressing challenges of deploying AI fashions at scale in a well-managed approach, enterprises must have a set of core capabilities in place.
This consists of the power to organize and curate excessive volumes of high-quality knowledge, have capabilities for observability, monitoring, benchmarking, and analytics, and be able to refine and enhance the fashions over time post-deployment, to call just a few.
We’ve discovered one of the best ways to handle these challenges is by having central platforms that may scale. In fact, all of this depends on sure precursors for achievement, just like the readiness of your tech stack, the power to belief and use your personal knowledge with a robust knowledge ecosystem, and high expertise. We’ve been investing in these foundations for years, which have ready us effectively for this AI second.
Q: How has the “data-as-a-product” method helped unlock new enterprise worth or innovation in your group?
We now have invested in constructing out a “data-as-a-product” method that curates our repeatedly used knowledge and knowledge that has the potential to drive innovation sooner or later.
Making that knowledge simply obtainable with a excessive stage of high quality is already rushing up our means to determine improvements and implement them, though we’re nonetheless comparatively early in our journey. We proceed to convey an increasing number of knowledge into this framework and champion the groups which might be constructing knowledge merchandise.
Q: What applied sciences or frameworks are most important for enabling scalable knowledge product administration?
One of many issues that’s most important for the success of information merchandise is elevating the position of the one that’s defining the intent and constructing the information product: the information product supervisor.
This can be a crucial, senior-level position that’s centered on not simply making knowledge standardized and accessible, but in addition excited about what the way forward for what potential makes use of of that knowledge is likely to be, and driving innovation from the information product itself. This additionally entails excited about the issues we aren’t doing right now, however which knowledge might allow, and serving to advocate for that throughout the group.
Q: In case you might automate one annoying a part of your day-to-day with AI, what would it not be?
For me, it comes right down to one thing extra mundane: on a regular basis dwelling life. It could undoubtedly be good to have AI autonomously remind me — and even assist me — with my checklist of home chores and ongoing upkeep in a proactive approach. I’d adore it if AI might supply trusted native and cost-effective service suppliers to assist with this work and schedule them in a approach that works with my schedule.
Leave a Reply