Video
PegaWorld iNspire 2023: Clean up in aisle 10: Dispute, exception, fraud, and sanction management in a disruptive payments world
As domestic and cross-border payment options for consumers, small businesses and corporates evolve and expand, how effective is your operation when there’s a problem? While straight-through-processing and real-time payments are the aspirations, they are not achievable in all instances.
In this session we will review strategies for successfully managing payment exceptions and errors in a world where regulation is trailing technology and a great client experience is demanded.
Transcript:
- This is "Cleanup in Aisle 10" and we had the longest title, the subtitle element of this. It actually didn't fit in the mobile app. They had to change the mobile app. So our title would actually fit here. So we broke the mobile app for the description of this and you're like, "Clean Up in Aisle 10." Why in the world did we name this Cleanup in Aisle 10? Well, when people say, what session are you hosting? I can just say, Clean Up in Aisle 10 and I don't have to read the whole thing and you'll remember it. So A, that's good. And it stands out for most everything else we do. Plus Kelly and I, we're bank ops people generally speaking and we clean up messes all the time. And I used to do it in my professional services career. So kind of felt it was apropo because in a lot of cases, those of you, how many people actively work in, it Better be everybody, actively works in the dispute area or supporting disputes? What are the rest of you doing here if you didn't raise your hand?
- [Kelly] Heckling, they're heckling.
- Yeah, they're heckling. They're here to heckle. So that's why we picked that particular title. Let me introduce myself. My name is Scott Andrick. I am an industry principal with Pega Systems. That's a nice fancy title that says I'm a payment geek, In this particular case. I've been doing payments, payment support in banking and in software for over 30 years. So when I started off, I didn't have as much gray around this section here and I didn't have two grandchildren. So I've been doing this in the domestic. I started off in banking in the domestic card market. I did work in international cards before I went to software and did work in professional services and in technical pre-sales. And have been running around doing dozens, maybe a hundred or so operational walkthroughs in this area. I spend a lot of time in this space and we kind of like. This is our collective internal, we're payment geeks. Don't call it tweedledum y tweedledee, 'cause that's not what we are. We're the payment geeks. That's what we do. So that's me and that's my side. I do mostly consumer, Kelly.
- Hello, I'm Kelly Wilson. I was in banking for 15 years before joining Pega. Focused on sales and trading, relationship management, also on payments. And I've been at Pega for about three years where I worked very closely on the commercial side of the house with our corporate and commercial clients who use us for payment exceptions.
- And so we're gonna split the consumer and the commercial aspects of this as we talk about payments. So let me get the clicker working. I thought it was working. Here we go.
- All right, so the world is changing, right? Payments are what drives the global economy. And right now there is so much change in the payments world. Actually the payment geek here sent me an article at 6:00 AM this morning about an extra talking about how do you improve your market infrastructure around realtime payments, which is very appropriate for today's, our discussion. So a lot of that change is positive. We've got ISO being rolled out for standard electronic messaging for various payments around the world. We have the Fed now is supposedly going live in July with realtime payments so that the US can catch up to a lot of other countries where there already is a realtime payments solution. Some of it's negative. There's a war started between Russia and the Ukraine and now that changes the entire global economy. And many countries can't do business with Russia. Sanctions them, they're not allowed to, or they are sanctioned, can't send SWIFT messages and everyone has to scramble to change who they do business with and have increased sanctions and flags for a period of time that's still ongoing.
- And then you've got a couple that just kind of, I get these all the time. They make you kinda scratch your head. So I'm just being bombarded right now. I don't know why it started all of a sudden, Venmo credit card from Visa. So we're taking a P2P and we're using old school visa credit card rails and we're combining it into some kind of new product. Okay, that's interesting. I don't know why I need both of those. And then I just actually this week I thought this was interesting. The government of Australia is eliminating checks as of 2030. They're gonna completely dismantle the ability to process checks. There will be no more paper in Australia after 2030. And my boss, our our boss DukPay based, gives us a hard time about the fact that the US still uses checks. By the way, that's C-H-E-C-K-S, checks in the United States. Yeah, but we understand that, that's a changing market. And there are market dynamics that are different in each location. But if we look at those drivers of change, we actually put them into three major buckets. And there's a proliferation element, a convergence, and a reemergence piece to this. This was all happening before COVID, but it certainly sped it up. Consumer choice, fintechs, government action. I have never seen this many options to pay for stuff and I've never seen the amount of change and the rate of change that we're experiencing today. And I've been at this for over 30 years. The challenge that we have with a couple of these though as we're rolling out all these new payments, and again that kind of goes back to the cleanup in aisle 10, everybody comes up with the bright, shiny new toy. They don't always think about what happens when it breaks. And that goes to the operations folks to fix, because some of the realtime payments are actually irrevocable. What are you gonna do when something goes wrong with something that you paid. I'm not talking fraud, I'm just talking about quality of good and services. What are you gonna do when there's no way to recover the money or they didn't really think it through. And it's gonna require a lot of manual effort to do that. That's a challenge in our industry for the bankers in our industry. You got a lot of convergence going on in terms of different payment types, ISO 20022 being the thing that's bringing to bear. One of our colleagues said it's like a language, but it's got regional dialects and variations. So it's almost like English, English and American English. So it's ISO, but it's ISO this flavor. So it's nice because there are some common elements, but it is not completely common everywhere. So we've got some work going on there. And then my next new favorite, the reemergence of point of sale financing, open banking, but also what was old is new again. Does anybody ever remember Christmas layaway or your parents? Could you raise your hand, be with me on this one. All right, buy now, pay later. Christmas layaway, right? It's just a new wrapper on what was already there. The thing that scares me a little bit about buy now, pay later, not only the credit underwriting element of it, 'cause that creates another problem, is actually the fact that people aren't using this just for large installment purchases. I've seen numerous articles from COVID through to just a week or two ago that talk about the fact that up to 20% of people are using this to buy groceries. Now tell me how are you gonna handle a quality of goods and services dispute on a buy now pay later when they didn't like the apple they purchased at your store. Not Apple as in Apple product. Apple as in I like honey crisp and it was mushy. You know, what are you gonna do? So operations folks got their hands full, Kelly.
- [Kelly] Yes, so along with all these new payment methods, the convergence, etcetera, there are a lot of challenges related to managing payment disputes, exceptions, fraud and sanctions. And we're gonna keep repeating that so we can say it as fast as possible. So Karin mentioned operational efficiency, fancy word, for do more with less, right? How do you balance rolling out that new technology. Understanding is it going to, how fast will it pick up? Adding technology to reduce manual work but also keeping the human element. It's a big challenge for all the banks. And whenever we talk to our banking clients, anyone in the industry knows there are so many different payment systems and like Scott mentioned, there's new ones that just keep coming. Regulatory compliance with every new product. There's a cost of compliance, a cost of regulation. You have to do right by what your government tells you. And if the government isn't telling you a clear answer, then you have to figure it out yourselves and do right by your clients. That has a cost which takes away the margins from those new products. And fraud and loss liability. During COVID, fraud increased in both retail and commercial banking. It actually increased significantly in check fraud in the US. Even though checks are diminishing, we have significant increase in check fraud. So there, how do you challenge that and put money into managing check fraud when it's something that's potentially going away. It's a struggle to figure out the right answer there. And then lost liability. Scott touched on in these real time payment world, there's no way to recall. Sometimes you get a warning that you shouldn't be sending that money to that person. Are you sure? But then if I do go ahead and send it, is it the bank's responsibility to fix my error or is it my responsibility? So those still, those questions aren't answered. And it's a question for Fed now too because if banks are going to use Fed now commercially, then there's a lot of amendments, recalls, et cetera, in the commercial world, how will that be managed going forward? And across all of this is customer satisfaction. That's what's most important, right? Making sure your client is happy, making sure their payment gets to where it's supposed to be and the amount that it is supposed to be. And if there is something wrong, then correctly supporting your customer quickly. But also balancing, again, whose responsibility is that? How is the bank protecting themselves? And also making sure the clients are happy. Because as we know, banking is a challenge and there are a lot of banks out there. And if your client isn't happy, then they can potentially cross the street and there will be another bank there ready to help them.
- And I want to really draw attention to this last point. And I know I'm ex operations, I get it. I was all about efficiency and being in regulatory compliance. I mean, and that's what I hear when I go talk to banks as well because that's what you're being measured on. There's somebody in your organization that also cares about the customer. There's a chief customer officer, there's somebody worried about customer experience. This is important. Lemme tell you why this is a moment of truth for your customers. It's interesting, BAI, again, I get these articles all the time and they said, you can't buy, you can buy new customers, you cannot buy loyalty. When your customer or your client has a problem in this arena, they are looking at you to make sure that they made the right brand decision. Full stop. Do you have their back? This is a moment of truth. Your ability to execute, whether you're at the front line in a digital channel or you're the person in the back office doing, say a chargeback or a SWIFT, you know, investigation, your ability to deliver on this has a material impact on the loyalty that your customers and clients view for your bank. And if don't believe that. There's a study that PWC did a few years before, it was right before the COVID pandemic and they talked about, you know, would you refer people and would you stay with you know, banks, you know, based upon service. But they asked a very specific question and I hung onto this one because I like to bring this up in this instance, they found that there was a 26% bump in loyalty if you took care of their fraud problem immediately. So if you have the appropriate systems and processes in place, because how many people have had a compromised account in the last 12 months? I'm actually surprised that the number of hands is that small. I was averaging a compromised account, one account a year for a while.
- Chase keeps texting me that someone's compromised my credit card, but it's from a 347 New York area code number. So I'm gonna guess it's not them.
- Yeah.
- App fraud in the UK is up 39%. The FTC said fraud the United States last year for all types of fraud, but most of that is financial up 30%. This problem is not going away folks. What's interesting, and I wanted to bring this out and I wanted to bring this story to life because sometimes if we're not customer facing, we forget this. So in the United States it's actually kind of global, but it's more United States based. There's an organization or a publication called the Financial Brand. And one of the editors there wrote about an experience that they had. They were going on a family vacation, they booked a hotel away, the family was going, there was this big thing going on in whatever the city was. And I keep going in and out. I might have to go to the other mic here. How many things can Scott handle at the same time? Dinner and a show? Here we go. So what's interesting is they show up at the hotel, it's 11 or 12 o'clock at night and the hotel doesn't have their reservation. Why? They booked it on a scam website. Now he knows he's gonna have a hard time finding a hotel. That's not where he went first. He didn't call a travel agent. Who'd he call first? His bank. And did he use the IVR or did he use a chat bot or anything else like that? Absolutely not. He says very specifically, and I wanna read this, that he bypassed all of the automation, all the things that you do for operational efficiency. He bypassed them all. And this is somebody who knows financial services too, by the way. He's an editor for magazine and he says that he wanted the uniquely human capacity to understand the context and urgency of my family's situation that would lead them to treat my demands with empathy and flexibility. I needed to know that I could trust my credit card provider to accurately process the details and reverse the fraudulent charge on my credit card. What if he had paid with Venmo or another real time payment that's irrevocable? How would you handle that? What's that gonna do to your loyalty? Because we've trained our customers over the last 30 or 40 years, when you have a dispute or a problem you call your bank, you have almost zero liability. Technically it's 50 bucks, but you basically have zero liability and the bank is gonna take care of it. Do we have that for all these new payment types? Nope. So you need to be ready for some change. What's going on in the commercial world?
- We're gonna do a little history lesson here. So looking back, I think you've all heard today, Pega was founded 40 years ago on the product of Smart Investigate, actually. I'm not sure if that was mentioned, but the 40 years was, and Citi was our first client, followed quickly by Bank of America and that was to manage payment exceptions, still is to manage SWIFT exceptions and investigations. And our top clients, we have 30 global banks using this product. Top clients get about 70% straight through processing with this product. So there's extreme benefit to keeping money moving, to stop it from being in suspense, to managing both investigations and exceptions, managing sanctions, RFIs, etcetera. But there's also a cost of compliance for us too. We work very closely with SWIFT. The Smart Investigate team partners with SWIFT to stay up-to-date with all their new functionality, whether it's gCASE being complimentary to GPI, rolling out ISO messages over the next, over the last few years, so that once that went live in March, Pega can manage both ISO messages and MT messages and we have interoperability between the two because that is crucial over the next three, two and a half, three years, while SWIFT accepts both message types because about 75% of banks in the world, as we know, as we've heard, are using MX now. But there's that 25% that are still on MT. So we also are staying up-to-date with what's changing in the world. And we're also thinking about how can we improve, how can smart investigate support real-time payments? We do that for a handful of clients and it's growing. How do we handle if this ISO messaging standard is standard, how do we extend to support other payment types such as ACH, where historically smart investigate didn't play as much. So there's that idea of also how are we growing with you as clients to make sure that we're providing the best service and that we are staying modernized with all the new payment types out there. Scott, do you wanna talk about Smart Disputes?
- Yeah, so started 40 years ago in the corporate side, fast follower. If you think about it, one of the most complex things that you can do in the back office is figure out how to do a charge back. So started with credit card disputes on the platform years ago. We've had a product called Smart Dispute in the market for over 20 years where we pre-configured all of those elements. And the idea here is to help with that end-to-end management and guidance to actually get you to a resolution. So there's automation built in there, you can do straight through processing, the compliance piece. Who are my US folks in the audience? Every time I walk into a bank, especially on the debit card side, they're like, help me with Rapyd compliance, right? Because the auditors are all over them. So how do we help you be in in compliance with your consumer regs. Visa, MasterCard, American Express, updating the rules every October and April. You gotta stay up on those and you know that was a great product. We've expanded that product. A year ago we introduced what we call our model workflow That was in smart dispute 8.7 last June. And we introduced that specifically to handle the kinds of expansion that we're talking about here in terms of payment types. So we're giving you a template that allows you to go out and tackle all these other different payment types and you can use that to launch into a unification strategy around how you do this. We've used that as our own development environment. So we've got Zelle, that was our first example, US based. We just in eight eight in March re-released ACH using the same model. And if go to our booth over in the innovation hub, if you haven't been there yet, they'll be able to show you a SEPA or a Fed Now example that they built in hours. The idea behind the model workflow is not for this to be a months and months kind of thing. It's intended to be a days and weeks kind of thing for you to be able to not only build, but also adapt for the market. Because at the end of the day, both Smart Dispute and Investigate are based on, you've heard this already at the keynote, the leading AI and workflow decisioning platform, low code. Remember that? All of the capabilities of that sit underneath here. But this is what's coming Kelly.
- [Kelly] So we've alluded to a lot of the different payment types, already mentioned some of them, but I thought it'd be helpful just to put them all up to look at them and think about what are those historic legacy payment types such as SWIFT, Visa. But now we have Visa Direct, legacy, credit cards, ACH, et cetera, and then all these new ones. Because every time I talk to our clients, I'm astounded at how many different payment systems there are. You know it, but you don't always think about it. So just looking at up here, all the different alternative real time payments now, maybe different blockchain systems that are potentially coming out. You've got groups of banks, consortiums working together to have trade finance solutions, to have distributed ledger solutions like Link, you might have central governments coming out with their own coins. I mean those are just five more that I mentioned right now that some are maybe live or in testing but not truly live. And then the embedded finance, like open banking and BNPL. It's a lot. It's a lot. It's a lot to say, it's a lot. So how are you going to be efficient about managing all these payment types, keeping up to date with the historic ones that aren't going away or going away slowly like checks and taking on the new ones.
- [Scott] And you heard Alan say this earlier, you know our tagline for over 20 years has been billed for change. Guys, the rate of change in payments is not gonna slow down. We anticipate that there's gonna be a massive proliferation and expansion for several years before there's really any meaningful consolidation on this. You not only have to be able to handle all the new stuff, but then all the changes when all the consolidation takes place. You have to have a system that allows you to build for those changes that are coming. Now, if you're thinking about doing this for the first time, as opposed to the RBC folks said, who's installed and who's looking at it? But if you're looking at this for the first time and you think, hey, I need to solve this problem, typically you're looking at a build versus buy solution. So do I wanna just build this on Java or do I want to go out and buy a packaged application? There's pros and cons to both of those. Obviously you control it if you build it in Java, but then you're also literally building from scratch. And for anybody who's had to do the Visa MasterCard updates every six months, I'm not really sure that's where you want to go. Also, it may take a little bit to get you to the point of some minimum viability for the product because you're starting from scratch. Packaged applications are nice, 'cause they solve the problem. The challenge that we find in a lot of cases is the packaged application is good for a rail or a small set of rails and doesn't have the flexible capability to keep bringing on new rails, new payment rails. Sorry. And ultimately it's not gonna give you the flexibility to change the product because they're looking at common denominator elements in terms of what they're gonna let you do. We believe that ultimately what you need is a platform for agility that is based on Pega that uses Smart Investigate and/or Smart Dispute, depending upon whether you're in corporate or consumer banking, because you're gonna get that ability to take a holistic approach to how you handle your exceptions. And I know some of you are shaking your head right now, it's like we're never gonna combine all these groups. I mean, Chris Thorpe who's sitting up here and I just did a session with ENMY a couple weeks ago and a webinar and literally we were talking about the fact that we used to laugh at people that said you were gonna consolidate payment rails and processing. That is not gonna be the case going forward guys. You have to realize that the silos are gonna come down. Now, I'm not expecting the corporate guys and the retail guys to get together anytime soon, but if we start standardizing on ISO, yeah, we're talking to different ballgame in terms of back office operations, reduce time to value in market, not just the first time out the gate, but all the changes that you need to make. And then also the ability for the operations folks and IT to work together in a codeless environment. The challenge that we've had for the past couple years is we've been attacking this in a very independent automation approach. We attack the channel, we attack the rail, we attack the consumer regulation, we attack little bits of this. We nibble around the edges and we end up with spaghetti. We've done automation. We've not done it in a coordinated and concerted fashion. What we're proposing here really is you should be looking at this as a unification of an exception backbone. That you build this commonly so that you can reuse this across channels so that you don't embed the logic and you don't do all the work multiple times because you've built it out for each channel. Online banking built their version, the web group built their version, the branch has their version, the contact center has their version. That's not efficient. Then you also have integrations into the back office. You wanna be able to use those systems multiple times, even if they are from different areas. Wrapper that with good case tracking, because these are long running processes. They can take months to ultimately resolve in some cases. That process orchestration and SLA management, absolutely key. You've got transparency, you know where the work is. You can break this work down as finite as you want to and give it to the appropriate knowledge workers. Or you can consolidate as a lot of people figured out during COVID and they couldn't move the work around and some groups got absolutely slammed because there was no way to move the work around. You need the ability to be flexible and adapt as both market conditions and unforeseen conditions come into play. How would this look on the commercial side?
- [Kelly] Here's another view. It's the same concept but a different view. You can build it up.
- [Scott] Yep.
- [Kelly] Think of it as a payment exceptions hub or payment exceptions event manager, where you have all these different groups that manage sanctions, that manage exceptions and investigations, that manage fraud and disputes and they can stay siloed if they need to stay siloed. We know fin crime sometimes has to be siloed for purposes of an anonymity, confidentiality, SARs, etcetera. But what you can do is as those cases are managed, you can push them to a central payments hub where internal teams can view them, where service teams could view them, where even the client could view the status of their payment. We talked to a lot of clients and commercial payments. 25 to 40% of the requests are, where's my payment? Where's my payment? Where's my payment? If your client had access to a hub, then they would be able to see it. And also you'll be able to improve internal efficiency because teams won't be chasing other teams internally. I used to be in the front office, I'd be chasing the back office. They didn't want to answer. Sometimes I'd go to their floor, they were trying to avoid me. So you can really help the entire firm by being more efficient, surfacing the results of status to a central hub, having the modern UI and UX, eventually consolidating systems as appropriate and when you're ready and basically create much better internal transparency and possibly surface it to the clients depending on where you wanna start and how you wanna create it.
- Now, no PegaWorld 23 presentation, we can't get through without talking about AI. I mean you got it in the keynote obviously. And the question is, how are we applying this within the the payments domain itself? And so process AI is one of the things that we've been working with here. And you can see from the chart, you know, different areas where you can start to bring AI to bear. Karim talked about a lot of the generative AI stuff today in his little go away slide, the process, the times they are AI changing, which was really interesting with the airplane taking off. But think about the practical elements in terms of where you could use artificial intelligence around predicting case outcomes, potential SLA misses, how long it might take to do something, assigning work to folks. Actually, 'cause I've been doing AI ML for about 15 years, how many people actually know what machine learning is? Just raise your hand real quick. Okay, so I'm gonna do the the quick primer on this and then we'll we'll flip over to some other elements. So normally if you think of a business rule, you think of something that we would call as deterministic if X then Y and it may be Z if it doesn't meet that condition. Low dollar write off. Great example. If it's under 25 bucks, write it off. If it's over 25 bucks, let it go. Machine learning is actually not that precise. It is actually a probability score. It is a prediction. It looks at the data that you provided and it's looking at that from a binary outcome perspective. And it's asking what is the likelihood that this thing that I'm checking on is going to happen? So you're gonna get a probability score. I learned this when I was at Cordiant, which is where we developed our customer resurgent hub technology, which is what Rob Walker's gonna talk about tomorrow. And the ability to present offers, cross-sell, upsell offers to people. We're now embedding this into the operational elements of the business as you can see here. So what we're doing is we're gonna give you yet another tool using process AI to go beyond just hard and fast business rules in areas where you want maybe a little bit more help. I think you've got a sanctions example for that, right?
- Yes, we have two clients using us, multiple clients using us for sanctions, but two specifically who have over 95% automation in the sanctions. Sanctions case management. Because using process AI, the ability to send and receive information automatically, the ability to look at historical data and predict that that is the same sanctions that we got yesterday, the same alert that we got yesterday, so we can discount it, someone doesn't have to review it again. So really high STP rates that clients can achieve by putting these pieces together of information that you already have, but making it work for you versus sending it to a human to work.
- And that actually was the next slide I just realized. We'll take that off. We're gonna take that off the blooper reel, but you were gonna talk about work distribution on this slide.
- Yes, so one feature of Pega also is work distribution. Scott just mentioned during COVID there were higher CS calls. Call centers were flooded for different requests to cancel flights, change things, etcetera. Market volatility means there might be floods in one area of trading on a trading floor and not as much work in another. So with workforce distribution, you can route work and monitor who has a lot of work and where people have less work and just automatically route that work to get everything done faster, keep everyone busy and without actually a manager having to manually look at that and be constantly adjusting the work. So you're saving the manager's time, you're getting the work done faster. There's a lot of benefits to that predictability of looking at the work and getting it to the right person.
- And one of the other ones, the second item that's on here is SLA adherence. This is actually the SLA rule is one of my favorite rules in the Pega system, bar none. Any of you have ever heard me present before, I usually stand up and down on the chair and jump around when I talk about this one because you can actually set goals and deadlines so that you're escalating the point of the work and you can actually have the system take action for you. You can automate the work based upon timers being fired. The way to take that to the next level with process AI though is actually asking the question early on, am I likely to even make the SLA? Maybe somebody is, maybe it's summer holidays and people are out and you're finding that they're not meeting them. Wouldn't it be better to know when the case is created that you're likely not gonna make the SLA and don't wait for the goal to come along to escalate, but to escalate even sooner. That's the kind of difference between existing rules and capabilities that you have and what process AI could do for you. One that we have, if I can get the clicker to work, here we go. One that we have over in the booth. I got Chris O'Leary here who's from our product management group. Over in the innovation hub, you're gonna see an opportunity. This is a proof of concept that we worked on where we're putting this into the product for our Q3 release of Smart Dispute. You're gonna do a chargeback, you've gathered all the data, you have a decision to make. Am I actually going to do the chargeback? Well, let's do some process AI. This is an $849 charge. What are the chances this is gonna get represented? And how confident are we about the represent aspects using process AI? Now what we did in this example that you see here on the screen is we preselected continue with the dispute because the likelihood of the resentment is low and we're still using a subject matter expert to make the decision. We could also automate this instead of it being an $849 item, maybe this was a $49 item. We're gonna work it because it was over a $25 load dollar write off threshold to begin with. But now what happens if it's a high likelihood of represent for 49 bucks, you've got a business decision, should I charge it back or not? You could actually automate not even sending this to a person if the probability is high enough and the dollar amount is over your low dollar threshold, but still under another level, you could completely automate this work completely. And I stole the sanctions example. We were supposed to talk about on this. Did you wanna talk about anything else relative to the process AI elements?
- Sure. Another one we hear a lot more about recently is trading exceptions. So a lot of times for anyone that works at global markets, the same trades with the same counterparties frequently fail. So what if you're using process AI to say, yes, those SSIs are wrong, or yes, this counterparty has a high likelihood to fail that trade because we're tracking, which trades with which counterparties fail. And we're gonna look at that data and potentially try to repair it or change it before the trade actually fails. When I was in markets, which was some time ago, the last three hours of the day, sales assistants are just running around fixing trades. You can automate that process to then use the data that you have to fix it.
- And so I wanna point out that not only Smart Dispute, but smart investigate are in the innovation hub. You're gonna be able to see our ability to do things like this, the process AI element plus other demos. If you've not been there, please go see it. Please go see the new model workflow elements that we've built in there and understand how we can help you in this arena. But there's more AI.
- [Kelly] There's more AI. So when we work with commercial clients in commercial payments groups, we regularly hear that 70 to 90% of inquiries come in via email. So another powerful element of AI is natural language processing. So what if you as an email came in, you could read and scan the email with OCR, natural language processing, and identify the intent, identify the client by the email that the client came from, identify account numbers, trade numbers, whatever's in that, and then automatically create a case and route that email to the operations team that needs to resolve it. In this case, it would be a trade operations team. You don't have emails sitting in group inbox, you don't have customers waiting for a reply. You can auto send a reply, we all get those, right? Thank you, we've received your reply and we'll reply within, we've received your email and we'll reply within 24 hours, four to eight hours. Now the client knows you got it. So this is a powerful way to create cases, route them to the groups that need to take care of them. And clients that are using this have achieved efficiency of 40 to 80% of their emails are going where they need to go without a human touching them. I want 40% of my emails to go somewhere magically so. But it's a huge powerful tool that learns quickly, is able to be trained based on incoming volumes, historical emails, etcetera, etcetera. And if you think about a commercial payments group where 40 to 60% of the requests are about payments, we've actually found from multiple clients across industries that have started using this tool, an interestingly similar statistic that 25 to 30% of the requests are duplicate requests. Whereas my payment, so you're auto routing a bulk of the information. Then when it's the duplicate request of you haven't, you haven't fixed it or resolved it, duplicate link Pega can identify that that has the same parameters and link the case to the same case. So now you don't have multiple operations people working the same case in different places and link those emails and again then escalate because now the client's followed up, it should be higher priority, your client's not happy, they're coming back for an answer, escalate it. If it can't be automated, it's already routed. But if it can't be automatically resolved, escalated to a human to then respond to the client. So there's a lot of powerful tools here in additional AI we want to focus on that could be really beneficial to those commercial groups where they're getting their inquiries and email. That. So we've talked a lot about building for change in payments. Pega has the ability to handle the complexity of your payments business. We are already managing most of our FS clients payment exceptions in either disputes or SWIFT exceptions or managing the handful in fraud and sanctions. So we wanna help you think about how the convergence and the new payment types that are coming and how we can help you create that payment hub or that payment backbone so that you are ready for the next new product that comes out, that you can unify your payment workflows wherever possible. You can keep them separate, as necessary, based on division or product type. Build layers of reuse. So you're ready for the next product type and it's easier and faster to be able to handle that new payment and unlock the AI and ML as we've given multiple examples of where you can get to high rates of straight through processing. Do you wanna go into a few examples Scott?
- And straight through processing is the nirvana. Then we go back to the four major elements. You know, number one, efficiency. Everybody's looking for straight through processing. Automation, definitely. Straight through processing is a risk tolerance question. It's a business question in a lot of instances. Are you happy with the data? Can you let the system go on business rules? And we have clients both in the consumer and in the commercial side that are seeing 70% plus in this arena. We have some people, we're still trying to figure it out completely, but they're claiming 80 and 90% STP. Now devil's in the details. Come talk to us, come talk to other clients. Actually come to this Smart Dispute. Meet up for lunch tomorrow. We had the Smart Investigate group today at lunch, which is where Kelly came racing over from. I'll be doing the same thing tomorrow for Smart Dispute. Come talk to your peers, find out how they're doing this. The other element that goes into this is also reducing the number of systems that are required to do the work, especially if you've got a knowledge worker that has to do the work. You're not gonna automate it. You could be looking at a dozen or more systems. This is an actual example from nationwide in the UK where they were doing, using 15 different systems. They dropped it to two. At the end of the day, if you can automate more things, especially straight through processing, and you can reduce the number of systems that are there, it should come as no shocker that you're gonna reduce the total elapsed time to get the work done. And this brings us full circle back again. That's what the customer cares about. They want their stuff fixed and they want it fixed fast. Your ability to tie this back to a moment of truth does still rely on your ability to deliver a great experience, but delivering it because you've done the automation work in order to be able to do that and to reduce your cycle times around this by as much as 85%. So with that, what we do, what we want you to do is we want you to go to the innovation hub. We absolutely want you to see the latest and greatest. Chris and his team have been killing themselves to get ready for this show. They've got great stuff to show, the latest things that are coming through, what we've just built and released and what's coming. You're gonna love it. We've got the Smart Dispute piece tomorrow and we have about five minutes for questions if anybody wants to try and play stump the chump. Nobody heckled us. Ooh, a hush fell over the crowd. Any question? Anybody got questions? We have microphones in the aisles. Just walk up. No. Either we did a really, really, really good job.
- Or they're tired after.
- Or they're tired.
- All right, well thank you.
- Thank you very much.
- [Kelly] Thanks for coming.
Tags
Ressources associées
![AI in Innovation](https://www.pega.com/sites/default/files/styles/1024/public/media/images/2023-11/ai-innovation-prev-card.png?itok=I5cCWnoL)
Exploitez la puissance de l'innovation par l'IA
Découvrez comment l'IA d'entreprise va contribuer à vos futurs résultats métier.
![Why Pega](https://www.pega.com/sites/default/files/styles/1024/public/media/images/2021-07/why-pega-preview-card.jpg?itok=ovPGRTVn)
Pourquoi Pega ?
Notre puissant logiciel n'est pas notre seul atout.