Welcome to FinTech TV.
I'm Marie Blair.
We're here at Money 2020, 2025 in Las Vegas, Nevada, and today I am joined by Toby Brown from Google Cloud.
So Toby, great to have you here.
Thank you so much for joining me.
Of course, great to be here, Remy.
Thanks so much for having me.
Well, a lot.
Happening as Money 2020 gets underway here in Las Vegas for this 2025 edition.
So first and foremost, I do want to hear about what agentic AI actually means for financial institutions.
Yeah, it's great to be here.
You know, it's always so exciting here in Money 2020.
You have everyone from the ecosystem all around and talking about uh what's happening and what's new, and there's just so much energy and it's great.
I think when we look at from the Google Cloud side what we're seeing with the financial institutions is that they're really looking at.
Agentic AI is an unlock for creating new capabilities that haven't been possible in the past and we really see that playing out across all sides of the industry, uh, so you look at like from a growth perspective as an example, one of the top topics I think around Money 2020 here agent to commerce, right?
So how are you using an ecosystem of agents to actually transact on behalf of customers.
Um, and certainly institutions is heavily invested in that.
We actually just announced a, uh, big partnership with PayPal this morning where they're using our agent to agent and agent to payments protocols to embed their own payments agent to help their customers and merchants transact.
And so we see a lot of institutions who are really looking at the distribution question, right?
So how are people going to manage and consume financial services in the future?
How can they build an agentic ecosystem that helps them thrive in that new world?
And from our side in Google Cloud we've been really focused on, uh, giving them an enterprise ready platform to do that, giving them open source protocols, uh, that help them stitch all that together.
And then you look at kind of more the defense of the operational efficiency side of the business.
It's also a huge unlock there, uh, where there's a ton of just complex workflows and financial institutions, I think previous ways of technology really haven't been able to automate.
You look at a space like lending operations, right?
So much unstructured data, uh, so many different edge cases, right, and variables, and so it's really hard, uh, to automate, but now we see a ton of progress using AI to unlock that, uh, automation.
We look at customers like Wells Fargo who have deployed uh Gemini Enterprise across their institution now and they're doing exactly that, right?
So they are thinking about how you can create these agents in these very focused ways to really tackle that end to end automation.
Yeah, and Toby, here at Money 2020, there are plenty of announcements taking place.
I'm glad you brought up that announcement that was just made here at the conference.
But of course when we're talking about financial institutions, the bottom line is key.
So tell us how.
The big banks are utilizing AI when it comes to driving ROI.
Yeah, it's a great question.
It's definitely top of mind for all the financial institutions.
We actually just did a big AI benchmarking survey a couple of months ago.
We hired an independent researcher to do it, and it was several 100 CXOs in the industry globally, and we asked them, uh, where are you seeing ROI and where you are seeing ROI?
What have been the biggest drivers of success?
Uh, and I honestly wasn't sure what we were gonna hear back from that because, you know, there's been more skepticism, I think, in the kind of media narrative recently, um, but what we heard back actually very much aligned with what I see in here every day with customers, which is that they are figuring it out.
They are scaling and operationalizing HIT AI, uh, specific to your question on use cases and where they're, where they're actually seeing that ROI hit, um, of course, the earliest area of success is really that low hanging fruit around improving employee productivity by improving operational efficiency.
Uh, specifically those were a lot of use cases like developer productivity, uh, like marketing content generation and campaign generation, uh, also customer service, another really popular one, right?
So using agents to, um, not only handle uh complex customer interactions in an AI only way, but also helping uh the human employees deal with more complex interactions in a more natural, elegant way.
And over time looking to now build that from a servicing and like perspective to a sales perspective and unlock growth from that as well.
Yeah, so I have a two pro question for you, Toby, because success is important, but also to be able to measure AI success, you need analytics and data.
So that's how this works.
Yeah, absolutely.
It's the old, you know, garbage in garbage out phenomenon.
So if you're going to maximize the value you get out of AI, you of course have to have a really high quality data foundation.
The signal coming out is only going to be as good as the quality of the data you put in.
Um, we've seen a couple interesting things on that front as financial institutions have again really worked to operationalize and scale the technology.
Uh, one is that it's a forcing mechanism for actually surfacing data quality issues that you may not know you have.
So in the contact center space is a great example where a lot of times they'll hook up the Gemini models to their knowledge management system, they turn it on and the output isn't what they expect.
It's not as accurate.
Uh, and they think that the models aren't working, but instead what happens is it's actually surfacing, you know, stale data or copies of data, things that they actually didn't even know were still out there.
So that way it's actually a great tool to help you accelerate cleaning up that data quality.
We also see a lot of people using Gemini to accelerate that, uh, that journey in a different ways like around data management and data governance.
So, uh, students spend an enormous amount of time and invest really heavily in.
Uh, improving things like metadata and data quality rules to just have higher quality data, and now they're starting to use Gemini to do that in a more automated way that really accelerates that cycle time, uh, for doing that work, which can be so slow and time intensive.
Yeah, and I think you mentioned a key word there, and that is governance, but also when we're talking about an industry such as financial services, we know that it is a highly regulated one.
So I do want to ask you how can organizations actually scale AI effectively and make sure they do it responsibly and safe.
Yeah, it's critical.
Financial services really is risk management at the end of the day, that's at the heart of the industry.
Everything financials do are looking through that risk-based lens first, so it's of course critical.
I think that uh We already mentioned one the first thing, which is of course the data you have to have that high quality data foundation in place.
Two other things come to mind that are also critical.
Uh, one is the talent side of the equation.
So of course there's a huge global arms race for AI talent.
Financial institutions are no exception, so everyone is thinking about their talent strategy, what they can do to build their.
Own internal uh talent pool and bench for AI as well as who are their key partners to help them do that.
Of course in the Google Cloud side we're, you know, shoulder to shoulder with our customers helping bring talent to the table and helping them solve that as well, deploying engineers, you know, right there to help them get up and moving and then doing the knowledge transfer and the training.
The third piece, which is also critical is the platform.
So, uh, you have to think about, you know, all the risk management frameworks that these financial institutions have.
Uh, and there's a lot of process and workflow and how that all occurs, and you know that to successfully get through that, you have to have a lot of things in place and so we're really focused from the Google Cloud side through our enterprise platform for AI called Vertex and building in all those end to end ML ops, all that work you have to do our model risk management.
And really helping all that the financial services ready out of the box so that really accelerates your time to production.
But thinking through that from a security lens, from a governance lens is absolutely critical to be successful.
Well, Toby, thank you so much for taking time out of your busy schedule here at Money 2020, and thank you so much for sharing all of your insights as well as your perspective.
Thank you, Remy.
I really appreciated the discussion.
Have a great rest of the conference.
Thank you.
