Name: Patrick Palomo | Smarsh
Uploaded: 2024-06-11T16:00:00.000Z
Duration: 13 min 59 s

Clips
News
More from AWS Financial Services Symposium

AI in financial communications: The Future of regulatory compliance and efficiency

In today’s technology-driven world, companies are making significant strides in transforming the industry using AI in financial communications. By changing how financial information is analyzed, communicated and utilized, AI is setting the stage for more efficient and intelligent financial operations, highlighting its pivotal role in the sector’s ongoing evolution, according to Patrick Palomo (pictured), vice president of solutions engineering at Smarsh Inc.Smarsh’s Patrick Palomo talks to theCUBE about AI in financial communications

play_circle_outline Challenges of technology adoption in the financial services industry

play_circle_outline The Crucial Intersection of Trust, Explainability, and Business Value in Regulatory-Grade AI Implementations

Info

Patrick Palomo | Smarsh

>> The AWS Financial Services Symposium, presented by theCUBE.

>> Good afternoon, AI fans, and welcome back to New York City. We're here at AWS Financial Services Symposium. My name's Savannah Peterson here for theCUBE, and very excited for our next chat with Patrick from Smarsh. Patrick, thank you so much for being here with me today.

>> Thank you so much for having me. It's been a great day. I'm excited to be here.

>> Yeah, I've been excited ever since our prep call last week. This can be awesome.

>> Likewise.

>> Just in case our audience has been living under a rock. What's Smarsh do?

>> Smarsh. Thank you for asking. Excited to talk about it. Smarsh specializes in managing digital communications. It just means emails, chats, phone calls, voice, WhatsApp, WeChat, whatever it might be, any way people communicate, we specialize in managing that, storing that, and we have a lot of fancy artificial intelligence that can help analyze that data.

>> What do companies do with that data?

>> So the vast majority of our customers are regulated financial services. So it starts with regulatory compliance for them. So they have regulatory record-keeping requirements, along with that comes things like internal investigations, litigation or e-discovery. And many of our customers actually have an obligation to monitor or conduct surveillance on their employees just to make sure they're not about to perpetrate the next major fraud or market abuse scandal.

>> Yeah. And I mean, you work with 95% of the top financial firms?

>> We do. We've actually been around for over 20 years.

>> Amazing.

>> We have over 6,500 customers today. They range from maybe your best friend starting an RIA in a strip mall from one to 10 people all the way to quite literally the largest banks in the world and everywhere in between.

>> Wow, that's exciting. I can imagine... Well, actually no. This is a question for you. Are the conversations that you're having with customers of those varying sizes, are they similarly excited and eager to jump on the AI train right now? Or is there a difference in scope?

>> It's hard. It hasn't everybody and their mother and their great-aunt heard about AI these days, so I think they are. It does vary based on the size of the institution. AI brings a number of efficiencies and risk mitigations, and I think the bigger you are, the more acute those problems are and the more willing you are to invest in those.

>> I think that's a great point. So do you actually think, and I say this with love to the financial services industry, not always known for their technology adoption, tenacity and the speed, I should say, at least because of the level of compliance and regulation around it, making sure if we're going to bring in a new tech that it's good, it feels like everything's moving a lot faster right now than it has in certain-

>> Very true. I think you see here today, we have a lot of firms talking about the amazing things that they've done just in the past two to three years, two to three months even. But the point you bring around the pace of adoption, that's a big part of what we do for our customers. So I think everyone wants to meet customers where they are. They want to communicate how they want to... How you communicate with your bank might be on an app or a chatbot or a phone call, whatever the case might be. So our CIOs, they want to make that easy for their customers to communicate with one another, and of course, to communicate with their customers, but they have compliance obligations. So that's where we come in, making sure they can stay compliant and still allow their teams to communicate the way that they expect to.

>> Yeah, absolutely. There's a fun term we discussed in our chat and in my notes, regulatory-grade AI. What is regulatory-grade AI?

>> We love to talk about regulatory-grade AI. So I think people appreciate tools being commercial-grade, or you might have surgical-grade steel. So our platform needs to be regulatory-grade, and what we mean by that, you need the resiliency scale and security, which is perfect since we're on AWS because they help support all of that. And when it comes to AI, it's really about explainability and defensibility. A pretty simple question, how do you know it's working right?

>> Right. How do you know it's working right?

>> That is the question. So it starts with the foundation of your training data. So far our customers, there's some use cases that might be great for a large language model that's been trained on the worldwide web or the internet, but for what we do, it's pretty important that that's really well controlled. You don't want to introduce bias, you can't have things like hallucinations. So it starts with the training data and it goes all the way from how you implement, how your internal teams review that, and more importantly, the reporting. That just allows you to explain that I know why and how this gave me the information it did.

>> How important is trust in that ecosystem?

>> I couldn't think of anything that could be more important. I don't know if folks appreciate the external scrutiny our customers face when answering that question, how do you know it's working? So maybe one practical example.

>> Please.

>> Let's say you have a million communications a day, which is actually most of our large customers probably have 10 times that. So let's say our AI models might spit just half a percent of that, honestly, in many cases, less to maybe just a quarter of a percent. That's 2,500 communications a day that a human being has to review, which might sound... I don't what your inbox looks like, but think of how many communications a human being is not reviewing. So how do you know what the machine is saying you don't need to worry about? So that's kind of how I describe thinking about regulatory grade. Answer that question, how do you prove it and how do you do it in an easy way with a nice shiny bow? So when someone comes and asks, you can explain that.

>> Yeah, I love that. I have 500,000 unread emails in my inbox, so that would be-

>> Just anything that's out from your boss, just delete it.

>> Right. Yeah. I can't even be bothered to take the time to delete it. I just leave it there. That's why they all keep my phone notifications, give people a heart attack when they look at them. I love that example. What are some of the use cases you're seeing produce results and add business value right away as you help your customers implement ?

>> So we have some classics. We've been doing this with AI for over 10 years. I know we're kind of talking about-

>> Amazing, look at you.

>> I know it's so topical today and ubiquitous the topic of AI, but we do it a long time for financial services. Our core use cases is our how do you make sure someone's not perpetrating the next Bernie Madoff's or Jeffrey Epstein type of scheme, the next LIBOR rate-rigging scandal. So that's the first place our customers look to us for. Tomorrow will be more interesting stuff, not just protecting against bad actors, but hopefully happy insights about your business and customers and revenue opportunities. But right now, it's absolutely about keeping capital markets safe and sound.

>> Yes. Well, it's important because within those capital markets, our individual capital as well, and it's all got to be safe.

>> I have a bit of an altruistic streak in me as well. It's I think protection of the everyday investor and integrity of capital markets. I think that's something that we can all get behind. Smarsh, as well as our customers, I think they all truly believe that we want a safe and trustworthy financial system.

>> Yeah, absolutely. The opposite of mean stocks all I'm thinking of right now. Yeah.

>> Absolutely.

>> We don't have to go down the GameStop rabbit hole, but easy to do. Where do you think customers are going to be with their... I mean, so, okay, wait, let me step back. So you've been doing this for 10 years. Did the phone just start ringing like crazy two years ago when the hype started to get more intense? Or were you already there with everybody coming on that journey?

>> This is how I would describe that journey. There was a wave of regulatory change post the financial crisis in 2008. I mentioned LIBOR rate-rigging or tax rate manipulation. So some of our larger wholesale banking institutions were dealing with the fallout of those crises pretty early, so many of these activities-

>> That makes sense.

>> Many of the activities were happening for a very long time, and that was the accelerant to really need to start expanding those compliance programs outside the US. The history of the US, the rules in this space go back a really, really long time ever since the advent of email. So in that context, it's been happening even longer.

>> What types of communications do you find are most risky?

>> Good question.

>> I'm curious.

>> These days, emails are one way people communicate and then that turned into maybe an AOL instant message, maybe once upon a time or a Bloomberg-

>> Oh, I remember.

>> Or a Bloomberg chat. And then today, if you think about how most people manage their communications at work, it's all over the place. It could be Slack, it could be Teams. I mentioned WhatsApp and encrypted chat channels, that's the most recent area that the industry has been focusing on. We call it off channel communications. So people are aware that they're being monitored, and they're looking to take things to another channel that might not be monitored. So most recently, about WhatsApp. And the next one is, of course, voice, just phone calls. People know that putting stuff in writing is not necessarily the smartest thing to do if you want to perpetuate-

>> Something nefarious....

>> a lawbreaking scheme. So voice is the next frontier there.

>> That is actually really interesting to think about. So just for the practical application of that, would you be listening in on their devices, on their edge device, their phone, or whatever it might be? Or how does that monitoring happen?

>> Many firms already are dealing with record keeping obligations, meaning just maintaining that record of phone calls with their customers. The CFTC regulations do that well as regulations in EMEA. So the next step is that we can take that process that and then run the same type of AI that could say, "Are they being secretive or they're perpetrating rumors or other types of fraud that could just flag a subset of those phone calls?" So human being can listen to them and say, "Hey, is there something happening here or not?"

>> Interesting. That is very interesting. What's the hardest part about your job?

>> Hardest part about my job. I think the hardest part about our jobs is we really share the press and the pressure of our customers. Most of the compliance officers we talk to, if you catch them in private, the weight of the world is on their shoulders. The preventing-

>> I haven't really thought about the pressure in fairness, but you're absolutely... I mean, point.

>> Talking about protecting the integrity of financial markets, protecting firms from severe reputational damage, and of course, just billions of dollars of monetary fines are also at stake as well, so there's pressure there. That's I think, the toughest part about our job, but we share that with our customers, right. We're doing it together.

>> I love that empathy. Would you say there's a culture of empathy when you think about that across the board at Smarsh?

>> So I was a customer for a long time before I joined Smarsh, so I've lived and breathed it for a long time.

>> That's cool.

>> But that's right. In all sincerity, there's not a version of this story where our Smarsh is successful and our customers aren't. We are symbiotic in that relationship that we both need to be successful in this scenario.

>> Absolutely. I would imagine you have a long standing relationship with these customers. I mean, they've probably been working with you for a long time.

>> We do. We do. Some of them I remember before I had children or before we were married, or oftentimes we work with customers at more than one firm, which is fun. It's a small world out there in financial services for sure.

>> It is a small world, but definitely a very connected world. How many people work at Smarsh?

>> Just over a thousand.

>> Oh, nice.

>> We're in eight different countries I think now. We've been growing a lot over the past 20 years from the shop that we once were in Portland, and we expect it to continue to grow.

>> I was going to say, do you think that genAI movement is going to end up making that scale a little faster? Because I can imagine you're really in demand.

>> It is funny I was just discussing with a colleague that, if I'm being honest, I think I underestimated the pace of how fast this has been happening, and I think that was ratified.

>> I think we all kind underestimate.

>> Today here, seeing what our customers are doing with investments in their own platforms combined with what we know we're doing with our customers, it seems like a rocket ship to the moon. Well beyond even what I would've foreseen 12 months ago.

>> Yes, it does. No, it's really... I feel like it's one of the cooler times to be in our world right now.

>> It's certainly not boring. The technology is amazing, and the way that it's been democratized is incredible. Everyone's got access to this cutting edge technology out there.

>> I know. It's just such a fun time. All right, kind of on that note, last question for you, Patrick, when we have your fabulous self on the show at this show next year, so 12 months from now, since we just mentioned 12 months, what do you hope to be able to say then that you can't yet say today?

>> No, that's great. Well, I already kind of teased that I expect voice to be the next frontier, so I would love to be able to come back here and talk about how many customers have expanded their compliance programs, identifying potentially risk behavior in phone calls across hopefully multiple jurisdictions and-

>> .

>> This is recorded, right?

>> Yeah.

>> At least 12 languages.

>> Oh, I love this. This is great. So we'll be able to play back this sizzle and have you back on and go through the whole thing.

>> All on record.

>> How old are your kids?

>> 16. So we got a driving test coming up.

>> Oh, exciting time.

>> I have a 14-year-old boy, and I also have a nine-year-old boy.

>> Oh, fun. What's the conversation like at the dinner table around AI?

>> AI is not AI to them. It's just the world that they live in.

>> It's just life. Oh, my gosh. It's just the intel inside already.

>> That's right.

>> That's really... Wow. Think about wow. Patrick, this has been so awesome. Thank you for taking the time.

>> Thank you so much for having me. This has been so much fun.

>> Yeah, this is a blast. One of my favorite interviews of the day, and thank all of you from tuning in wherever you might be on this beautiful rock. We're in New York City at AWS Financial Services Symposium. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.