GenAI: The Superhero in Disguise with John Kain

This is Techstrong tv. Hi everyone. Welcome back here to Techstrong tv.

You know, AWS reinvent will be something we're all talking about for the next couple weeks heading in, well, we're in the holiday season, but heading towards Christmas and New Year's. Um, it's kind of the last big event of the year, and so I guess it's appropriate that we have someone from AWS on Tech drunk TV with us today. Let me introduce you to John Kane.

John is the head of financial services market development at AWS and that's a mouthful. We're gonna, we're gonna hear about that from John, but we're gonna hear a little bit about kind of John's own journey as well. So let's welcome John.

Welcome to Text Drunk tv. It's great to have you on here. Thank you so much, Alec.

It's pleasure. So John, we're gonna get into the financial services market development, but let's hear a little bit about you and your kind of road. How did, how did you come here today?

Well, so I've been at j at AWS for a little over almost eight years actually now. Um, fantastic. And before that, really 25 years from the financial services industry, uh, prior to coming to a WSI in a compliance role at JP Morgan, was actually a lot of fun.

Was building out a lot of the market surveillance systems, looking at jps trading activity across multiple asset classes. For that I ran a risk management business at nasdaq where was looking at real time risk across markets. Did a few years in private equity, investing in early stage startups, but really grew up as a technologist within the industry, starting on the corporate lending side at Citi, and then sort of over the counter trading at Bear Stearns.

So I've had the pleasure to work in, Well, that's the name I haven't heard. Yeah, exactly. It's been a while, right?

Mm-Hmm. Uh, but you know, that set of experiences from about the technology, a business and investing perspective, as well as what it means to operate in a regulated environment really was sort of the ideal background to what I actually do here now at AWS. And that's a great segue.

What is it you do now at a WSI Have the pleasure of having a team that works across our customers globally, helping them leverage AWS very much from a financial services context. And now there's no one financial services industry. So the team is organized into our industry segments.

So the team that covers banking, insurance, payments, capital markets. And what we're trying to help our customers do is actually drive business outcomes that are relevant to the financial services industry. So much more focused on how we improve any money laundering or how can we actually improve trading strategies, how can we improve claims processing the actual or processing insurance?

And then how do we help our customers leverage those AWS services that will help them achieve those goals. Secondly, we realize that AWS doesn't offer everything our customers needs, particularly from a financial services perspective. So work very closely with the AWS partner team to make sure we're bringing the right community of both system integrators and ISVs into the AWS um, partner network to make sure our customers have that choice to build versus buy, and also work very closely with our services team, helping them understand sort of the needs of financial services customers so we can influence AWS's roadmap to better serve the industry.

Love it. There was a lot packed in there, John. We are gonna have to peel it back one layer at a time, like an onion here, but we're, We're busy here at AWS Alan Ain.

It's the truth. It, it's certainly is, right? Um, it sometimes boggles the mind when you look at the many, because I mean, here at Techstrong, for instance, right?

We deal with a lot of the AWS partner, uh, organizations, and I know, I mean, there are so many silos, and I don't mean that in a bad way, but so many different verticals that deal in the partners that it, you know, we're a small team, it consumes us, right? Trying to deal with all of it. So AWS is is certainly a very busy place.

Well, What I think is exciting about that, Alan, is like from a partner perspective, we're increasingly seeing our partners deliver financial services a little bit like we deliver technology sort of via API on demand at a global scale. So increasingly that partner element to sort of what AWS brings to the table allows our customers just to develop financial services applications ever more quickly. You know what, and that hawks back to the original AWS pitch, right?

Which was deliver like Amazon, sell like Amazon, right? If you wanted to have that Amazon, like scalability, intelligence and everything else, use a this goes, I mean, I'm old. Oh, Absolutely.

You know, and that's when you first started seeing AWS out there, right? Amazon like infrastructure. Now of course everyone today talks gen ai.

If it's not gen ai, it's gen ai. And if it's not gen ai, it's still gen ai, ai. But, and, and it's funny, the thing about gen ai, John, you know, it's such a double-headed, or two-edged sword, double-headed coin.

There are things that gen ai, gen AI does and will do for us that's gonna revolutionize not only how the financial services market operates, but our day-to-day lives. So Alan, I think even before generative ai, I think we have to realize that artificial intelligence and machine learning technologies have had such an incredible impact on the industry's ability to kind of detect and reduce the instance of fraud, you know, across the financial services industry. Um, you can look at firms like MasterCard who are using kind of machine learning technologies to look at payments transactions.

And by using just deep learning and other machine learning techniques against those patterns, they're able to detect three times as much fraud, but also reduce the amount of false positives that ruin the customer experience by like a factor of 10. So that ability to take machine learning and actually use it to reduce the disruption in the financial industry has been going on for quite a while. And I think sometimes you forget like how detailed that can be.

Um, new data who is acquired by MasterCard, they were actually using biometric signatures on your phone, the way you hold it, how fast you type, how you to actually create a signature of yourself to reduce the ability of fraudsters to actually take over your device and pretend to be you. So we've already seen those technologies being effectively used to help make it harder to create fraud. I think the areas you're talking about probably are the new generative ais and the areas of scams, right?

Where you actually sort of enable, um, a bad actor to try to use the latest technologies in a way to trick people. But we're seeing that those same type of technologies can actually be used to detect those patterns as well. And increasingly what we're seeing is that, uh, the ability of our customers both to leverage generative AI to understand those conversations better, but also to mix them in with more traditional technologies like graph databases that allow you to see the relationships between people and entities give you that insight beyond what sort of an individual bad actor can do to kind of help or detect and reduce fraud.

Absolutely. Um, again, not, I don't wanna get, you know, too under the covers, but how is is AWS offering that as a service, if you will, to financial services, uh, partners, customers? Yeah, Alan, what we're doing is making sure that we're meet matching the needs of the industry from a technology perspective and allowing our partners in the industry to bring sort of that last mile of expertise, particularly when it comes to data to bring insights into what they know about their customers in a much more effective way.

And so at Money 2020, just a few weeks ago, we were demonstrating with Nvidia how we could use the latest kind of technologies leveraging sort of that same underlying technology you use to build foundational models to actually build fraud models, but to do it at a scale that's ever bigger and see new insights that you shouldn't, couldn't see before. But the real experts at being able to do that are our customers who have that relationship with the customer and have that history of data. And we look at what really differentiates firms in the financial services industry, it's the data they have and how they can manage it.

Now, certainly we will kind of raise the base of what's available from a technology perspective. So we have, you know, fraud models that are kind of trained, that are shared out, um, within our kind of stage maker, which is our machine learning platform that gives our customers a jumpstart on kind of getting there faster. But the real expertise and the real ability to do this is our customers and partners that actually work in the field every day.

You know, I was talking to someone in the gen a gen AI space last week, John, and, and it's in line with what you said. What what amazed is them is, you know, they produce, in this case it was hardware platforms, right? And it was called, uh, uh, GPU as a service, let's call it.

Right? Um, but what amazes them is what their customers do with it, because that is where the creative process really kicks in, right? Because your one organization as big as AWS is it's one organization, but when you have thousands of different organizations pushing, pulling, stretching, and you know, based upon their own needs and wants and, and ingenuity, it's like a, a giant Petri dish, right?

You never know. You don't know what's going to turn out there. Here's another thing though, I see, and it, and it kind of mimics the whole AWS cloud experience, right?

I mean, AWS originally infrastructure as a service, right? I, I give you your hardware platform hypervisor and you build os application, have at it, right? Same kind of thing in Gen ai.

If you, all you're looking for is that basic infrastructure, GPU type of infrastructure and, and what you build and you put your apps on there, fine. What, what I suspect is AWS, and, and again I'm sure we'll hear a lot of it at reinvent, is it's more than just that GPU infrastructure layer. We're starting to give you the building blocks, whether it's APIs or full-blown applications using gen AI that will allow you to build your business and take it in ways, you know, that we didn't really contemplate right.

To. Yeah. Where is Ws should that Sure.

Um, so we think about sort of generative AI and AI technologies in general and sort of layers, right? And some of it as you mentioned, is that just core infrastructure layer, making sure we're giving you the most performant, most secure infrastructure to build AI applications. On top of that, we have our kind of AI framework layer for people building their own models, something we call SageMaker, which is our machine learning platform.

But in the generative AI world, it's really Amazon bedrock, which allows our customers to use between different generative AI models and bring them in and choose the one that best fits their need. But increasingly we're seeing on top of that the need to bring generative AI into applications themselves. So with some of our first party products like Amazon Connect, which is our call center product, we have the Q family, so Amazon Q4 Connect, which actually brings generative AI technologies into our applications.

It's the same type of thing you'll see in the industry. Um, so Smarsh is one of our partners in the communications space, and they built an intelligent, uh, agent system that's embedded within their platform. So when you think about their compliance workflows, when they do communications, uh, surveillance, you don't ever actually have to exit the platform to get the benefits of generative ai.

They've built generative AI into the workflow, so they take advantage of both the context of what's going on from a compliance perspective as well as the data on their platform. Now underneath that's powered by AWS technology, but within the context of the application, it's right there for you and embedded. So you don't ever have to kind of leave the application to get the benefits of generative AI technology and increasingly both in our own products, but in the products of our partners.

We think you're gonna continue to see that generative AI integration within each of those workflows across the industry. Agreed. Agreed.

Let's, John, if you don't mind, let's focus right into financial market vertical, though that, 'cause that's your vertical. I don't wanna leave people with the impression that we're only using generative AI and AI technologies to fight crime or, or mo you know, uh, bad actors. Let, give us, leave us with some positives here in how we're using gen AI just to make the financial services, uh, delivery of, of, of solutions better for, for all of us.

Sure. Yeah. We're seeing gen ai these in a number of different areas, Alan, the, even on the areas we're talking about on the, the financial crime side, we're seeing huge time savings from an investigation perspective.

A lot of the output of what we've been talking about is like, oh, if you see a bad act, you have to actually investigate it and sort of see what's going on. It's one of the areas where generative AI technology takes away sort of that undifferentiated heavy lift for investigations. Like in the financial crime area, if you're looking at patterns of behavior and you see an alert, somebody's gotta go investigate and figure out if that's really like potentially criminal activity or if that was just a false positive.

Um, NASDAQ's Ferin is an a, uh, does any money laundering platforms, they've actually used derive AI technology to automate the aggregation of information that goes into that investigation. Helping investigators make the determination, whether it's an investigation should be a criminal one or if it's a false positive, they've been able to save over 80% of the time that goes into that human investigation just through the use of generative AI tools. We see it in the use of the market area where NatWest is actually doing more targeted communications to its customers and they're finding by using generative AI to make personalized communications to customers about products they really need.

They're seeing like click-through rates of like four times, you know, and orders of magnitude increase in sort of the conversion rates for certain products. Um, from a customer experience perspective, which is probably what some of the areas I'm most excited about, um, we see the ability at, particularly with call centers and interactions with customers, the ability is generative AI to take the conversation. Let's say you and I were having transcribe it and summarize it.

Now that's a time savings for like operators to the call center. But what's more importantly, it gives them insights into why their customers are calling, how they're doing with their customers, and how they can better serve them. So you can go and now take those summaries of tens of thousands or hundreds of thousands of customer interactions and query it as like, why are my customers calling me?

How am I serving them best so I can change the user experience and actually improve the way we're delivering financial services to our customers? So that ability to use ative AI technologies in much more unstructured ways, particularly in the communication area, I think we're seeing tremendous benefits and I think we've just scratched the surface. Oh, ly, You know, and I have this three from now, we're gonna be talking about, uh, reinvent Alan that last year we had done a presentation with three of our customers who had just started putting applications in production last year.

And I think one of the exciting things is we're gonna revisit that sort of session again this year where you'll see the increase in complexity of the applications. Like people are doing much more sophisticated work with generative ai, taking larger problems, breaking down into smaller ones using generative AI agents and sort of actually driving the way that the industry's able to conquer even more complex tasks even just from a year ago. So very exciting time to kind of be, and like you said, it's still the very beginning of generative AI adoption.

Absolutely. John, if someone, maybe they work in the financial services market sector, where can, is there a specific place within AWS they could go to, to get information on, on this particular vertical? Yeah, from financial services, if you just go to the AW S website and we actually have a financial services section, it'll kind of give you the breakdown of what we're doing from an industry segment perspective, as well as some of our more interesting customers, um, as well as the link to what's going on at reinvent this year.

And if your customers, uh, are people on the line haven't actually looked, there's a financial services guide for reinvent, which actually breaks down what's gonna happen over the next week, day by day into where our financial services customers showcasing how they're using AWS to drive innovation in the industry, as well as some of sort of the key use cases in the industry that we think are most exciting. I love it. John, thanks so much for coming here on Tech Junk TV and talking a little bit with us.

Please do come back if you have an open invitation anytime you'd like. It's a pleasure to have you on, Alan, thank you so much for the time. I appreciate it.

Alright. John Kane, head of financial services market development at AWS here on Tech Drunk tv. We're gonna take a break.

We'll be right back.