The Collections AI Ecosystem: What Leaders Are Doing Differently

Adam Parks (00:01)

Hello everybody, Adam Parks here with a receivables LinkedIn live webinar. Today we're gonna have a really interesting conversation around artificial intelligence ecosystems and trying to better understand how we talk a lot about Opti-channel and Omni-channel, but how are we gonna start actually coordinating these tool sets across these various channels and using all of the data at hand to drive performance.

Matt Maloney (00:20)

Thank you.

Adam Parks (00:25)

And the reason I thought today was such an interesting conversation for me is because we have so many different perspectives joining us today. We have Shantanu who's joining us from Prodigal who actually has been building these tools and was one of those front runners in artificial intelligence within the debt collection world. We've got Kyle Cohen, who is an attorney law firm working multiple states and Matt Maloney who with First Financial Asset Management who also provides us with the perspective of the debt buyer in agency and It's I never thought I was we were joking about this right before we went live but I never thought if you told me five years ago I was gonna be having conversations with attorneys that were actively using artificial intelligence tools to collect debt I would have told you you were crazy and so I'm really excited to participate in these conversations because it clearly the the landscape has changed and for a variety of reasons that we can start to talk about but I think we're in a very interesting new frontier when it comes to the deck collection industry. And so gentlemen, for anyone who has not been as lucky as me to become all of your friends through the years, Shantanu starting with you, could you give us a little background?

Shantanu Gangal (01:33)

Absolutely. First and foremost, Adam, thanks for having me on this and always exciting to brainstorm with you, Matt, Kyle. We've had the pleasure of working with all of you for several years now and it is exciting to kind of get the group back together. In terms of my background, I have been doing consumer finance for about 15 years at this point and I've seen some of the industry change very rapidly. And so right when we were starting Prodigal, we kind of were at an inflection point of maybe three macro trends. The first macro trend was the change in the consumer ecosystem, lot of like, for example, Binopulator was starting to take off, credit sizes were changing. The second is kind of the regulatory situation with like the CFPB having gone through their first wave of changes, but at the same time were about to introduce RegF 7 and 7 and so on and so forth. So that's the second macro trend. But most interestingly was the third trend, is around AI, which is my background where we were very clearly seeing that AI had disrupted several industries and right in 2017, the current paper which has effectively like given a lot of flourish and impetus to the current AI boom was published in 2017 and both me and my co-founder kind of looked at it, read it had a literally like mind's blown and we started to do this, which is what we've been doing for the last seven years, kind of bringing AI to understand consumers better. And most importantly, do it in a way that is most completely in line with the current regulation and creates a win-win-win. It creates a win for the consumer because they get a better experience, a more convenient experience. It creates a win for the lenders, loan servicing and collection agencies because it increases their profit pools and it creates a win for us because we are able to service that technology. And so all in all, think that kind of trifecta of good three macro trends led to the birth of what we do. That's AI agents for loan servicing and collections. And a few years later, here we are.

Adam Parks (03:47)

You foreshadowed a whole bunch of things that we're going to talk about in this conversation. Matt, going on to you, could you tell everyone a little bit about yourself? ⁓

Matt Maloney (03:53)

Sure, yeah. Hey, my name is Matt Maloney. I am one of the co-founders and CEO of First Financial Asset Management and the FFAM 360 Alliance of Companies. You know, background on me, I was, interestingly enough, one of those individuals who was born into this industry. I've been around it my entire life. I remember being in high school working in the industry as a collector in the summer days collecting on Bell South advertising and publishing YellowPage accounts. ⁓ So I've gotten an interesting look and I said, you know, I don't think I want to get into the industry. I want to do something different, which I did initially coming out of school.

Adam Parks (04:23)

I'm

Matt Maloney (04:31)

did something different in a totally non-related industry. But I've been in the industry as a owner operator for 25 plus years. And much like Shantanu I don't know how many rounds or phases of kind of what I call macro operational changes, but I've seen a lot of change over time and really have the benefit of seeing it from the inside out as we've shifted to where we're at today. Super excited about where we're heading as just overall in the new industrial revolution, the artificial intelligence revolution of where it's going to fundamentally transform the industry, what's already happened and where it's moving towards. So very excited about that and happy to be here. Thanks Adam for having me.

Adam Parks (05:14)

Thinking about it like the Industrial Revolution is very interesting. I want to come back to that. But Kyle, could you tell everyone a little about yourself as well?

Matt Maloney (05:18)

Yeah.

Kyle Cohen (05:23)

Hello, my name is Kyle Cohen. I am managing partner of Cohen and Cohen Law. It is a law firm, a collection law firm located out of New York City. And a little bit about my background is I was born into this as well. I came into a family law firm, took it over a couple of years ago. And the way I'm looking at this is a bit different than some other people is, as everybody knows, it's no secret that law firms are historically behind on the times when it comes to technology, where they focus their time, resources, energy. So I come into this as a law firm that was, like many law firms, not technology focused. That's not a lawyer's background, typically.

And the people running a law firms, that's not their back. I looked at this as how can I kind of rethink the law firm model, what it is that we do as a law firm, what technology is out there, how can we be cutting edge, how can it be beneficial, what is it that we want to do and what is it that we want to not do and avoid and make sure that we're still doing our jobs as attorneys. So I think I come at this from perspective

Matt Maloney (06:06)

No

Kyle Cohen (06:33)

of lot of people who haven't been technology focused and are now seeing the way the world is moving and how can they kind of catch up and get ahead and use technology in a way that they've never used it before.

Adam Parks (06:46)

Very interesting and I want to go back for a second and talk about the industrial revolution piece because I think that's really interesting. We are kind of in a new industrial revolution that's happening. It's going to change the way in which we educate our children. It's changing the way right now in how we do business. It's changing the expectations of the consumer themselves.

Similar to what happened, the lifestyles have changed and expectations of the general public change and I think we're gonna be witnessing that over the coming years here as well. If we're not already seeing that in terms of, we've had conversations about the shame factor and how consumers may just want to self-service so that they don't have to engage with somebody else and then they can talk about generational changes and all of that. But...

Matt Maloney (07:23)

you

Adam Parks (07:34)

Before we jump into all of that, Kyle and Matt, both of you have gone on an AI journey as a professional debt collector in your respective disciplines. How did you look at that evolution of, I need to, what was your trigger that told you that you need to start making these technology investments and you need to really start moving down the path of leveraging these tools?

Matt Maloney (07:55)

Yeah, I can take that first. My journey in this area really dates back to 2018. I distinctly remember we were looking at, how we were doing things, what was kind of from just an overall, and put artificial intelligence aside for a second, just look at the digital transformation that the industry was starting to adapt to, starting to accept. I mean, this is prior to Reg F, which occurred in 2021, where we were looking at how we were engaging with customers, how consumers wanted to be engaged, and we kind of took a little bit of a leap, from our perspective, a little bit of a scary leap of faith and started communicating with the customer via email, started communicating with the customer via SMS message. We started overlaying our, what I would call our version one of artificial intelligence where we utilized an AI rooted system for speech analytics to sit over and listen to all of our phone calls and record all of our phone calls and then score all of our phone calls and flag calls for issues and compliance so that when we first, you know, kind of approached it, we looked at it from a compliance perspective on the artificial intelligence side. But overall, for lack of a better description, our digital transformation, you know, started in 2018 and then kind of evolved from there. And so as we got more more comfortable with how we were engaging with the customer digitally, and then we got more and more comfortable with the accuracy of how we were using artificial intelligence not to engage with the consumer at the time in the use of AI. It was more back office stuff, like I said, on the compliance side. But as we got more more comfortable with that, it just was a natural evolution for us. then it brought us to where we're at today, which I'll talk about throughout the call, throughout the webinar today about how we're utilizing it, where we're utilizing it. But again, our journey started in 2018, and we actually went specifically to a client of ours and we ask for their permission to start using these digital channels and then eventually that morphed into being able to use and implement the use of artificial intelligence into that process as time went on.

Adam Parks (10:10)

How tightly tied was that AI process to the digital transformation?

Matt Maloney (10:14)

Yeah, well, I think that obviously digital the digital piece was and the non-AI piece was adopted quicker across the operation when it comes to the consumer touch point. AI was a lot, we slow rolled it a lot more. So again, we were using the artificial intelligence component behind the scenes, again, more for compliance or operational effectiveness or efficiency. And then, we slowly rolled it out to our consumer facing engagement.

Obviously, mean, they were right there along each other. Maybe I would use the analogy they were both trains, but maybe just on two different tracks slowly going along at slightly different paces.

Adam Parks (10:51)

Okay. And somewhat dependent on each other. Kyle, how about from your perspective, looking at it from the law firm's seat, what did your journey look like to, what was your catalyst to making the investment into artificial intelligence tools?

Kyle Cohen (11:06)

I would say that there's two things mainly. The first being is as a law firm and similar to probably collection agencies, we all kind of do the same exact thing, right? We are retained to collect money. So the question is, is what can be your differentiator? What can give you that edge? So when it came to me personally and my skill set, my background, I knew that I can have an edge on the technology side, on the digital outbound side, on the artificial intelligence side. And that's where I can, you know, I might struggle a bit more to compete with people who have been doing call and collect for decades. And I don't, you know, I'm younger. I don't have decades of call and collect experience or running that. But what I can do is look at this brand new technology, get in early, put just lot of research, a lot of conversations, a lot of meetings with potential vendors and compete that way. And the second way I'm looking at things is similar to the industrial revolution aspect of it is I like to think what are things going to look like in 10 years? And this is my guess in that 10 years there's going to be less more profitable businesses utilizing artificial intelligence more than we've ever imagined. And those companies that are doing that in 10 years are going to be the ones who are right now, today, making those investments. And, you know, maybe I'm wrong, but if, you know, if I'm wrong, you know, doesn't hurt me. If I'm right and I'm not making those investments or other people aren't making those investments, I don't think they're going to be in business in 10 years. So I want it to get started now, today, where even if everything isn't perfect, when the technology progresses, I'll be on the cutting edge of it rather than starting from zero.

Matt Maloney (13:00)

Yeah.

Adam Parks (13:01)

So that true competitive advantage in a space that's highly competitive, being able to leverage that, I think makes a lot of sense. I expected to hear, I don't know from who, but I expected to hear a little bit more about the current state of the consumer. in my mind, and when I looked at these artificial intelligence investments, that was the first thing that came to my mind. And Kyle, I like your approach of like, want to be first to the market. Like the world is going to change. Like let me be on the cusp of that wave.

In my mind, I start looking at, three, four, five years down the road, I'm going to have to do more with less. I'm going to be missing. These consumers are going to get crunched. And what we saw in the debt collection industry survey this year was expectations. People have experienced rising, rising volumes. They're expecting to see more rise in volumes, but at the same time, we're seeing challenges to liquidity and then you start stacking in student loans and other things, these consumers are very stressed. We've seen over the past, let's call it 10 years, a move towards self-service technology. And we've definitely seen more comfort and confidence within self-service technology. We've also seen more comfort and confidence in regard to subscription services. And I think the consumers are becoming more comfortable with those recurring payments because you didn't have a subscription at Blockbuster, I guess maybe you did. But you were happy to pay for a subscription for Netflix now and over that time period is kind of how that customer has changed. Are you guys seeing the impact of those changes now that stress on the consumer and do you as that changes are you seeing any are you seeing any change in how they're engaging with those AI driven channels? Is there a comfort level that's starting to grow?

Matt Maloney (14:43)

Yeah, well, and you know, I think going back even, Adam, and you used to hear, you know, maybe seven, eight, nine for us, was kind of that's that was the aha moment for us back in 2018 is this consumerism approach, right? We're in a consumer, there's consumerism coming to the accounts receivables management debt collection business. And, and so just like with most industry that you see in the US, we're really no different. Follow the preference of the consumer. Like the consumer is going to drive our decision making and how we operate our business. And I think that artificial intelligence isn't just changing how we talk to consumers, it's transforming everything that we do from the inside out, right? Both on the front end and the back end. And so, your question about are we seeing the consumer help us drive that decision to move towards the utilization of artificial intelligence and whether it's automated communications via SMS, MMS, RCS, however we're communicating with them, or whether it's, you know, your chat bot, your conversational AI, as I like to call it, whatever it might be, without a doubt, I mean, there's no question that today, the reality of it is, that consumers are answering the phone less and less. So traditional phone strategies are starting to wane a little bit, right? And they want to communicate more and more, you know, through this, right? Through this thing here. And some of that communication preferences might be through, again, SMS or live chat. Some of that might be through email. But some of it, especially dealing in debt collection, where there's maybe a hint of, you know, maybe a hint of shame in terms of like they don't want to deal with it or just disdain like I don't want to deal with this. It pisses me off. Yeah, I may have gotten this scenario. I know I the bill whatever happened. Sometimes they don't want to deal with the human and so we're starting to see the communication adoption rate increase literally, I mean monthly we're seeing it monthly in our organization in terms of them their willingness to communicate with conversational AI right with a chat bot. And I mean even today just before this call I was looking to see how much we had collected with our conversational AI today in communications where the consumer made the inbound phone call, talked to, our IVR is gone. It's now, for lack of a better description, an intelligent virtual agent talking to the consumer. And then that first point of contact is that intelligent virtual agent or that conversational AI. If we can resolve the account, if that consumer prefers to resolve the account that way, then they'll resolve it.

But if they want to go to a human being, can opt to go to a human being and talk to a human agent. But yeah, the consumer is driving our decision to move in that direction because ultimately our business is based on communicating and contacting and having right party contact with the consumer to engage them for payment.

Shantanu Gangal (17:38)

Yeah, I strongly concur and maybe like building upon the kind of industrial age, like when cars came about, not only did they replace horses, but people actually put like a person to drive the carriage and they were like, no, no one drove it themselves. There's someone else who was supposed to be specialized doing it. But very quickly people realized that you want to be in control and you drive at your own time, you kind of control the whole thing, you take out the middleman. I think that experience has become better. So as you kind of get closer and closer to seeing technology firsthand, people suddenly like start seeing new things. And no doubt with the launch of like chat GPT and a lot of these like modern interfaces.

Everyone has had some experience using like a Gemini or a chat GPT. And so that has definitely helped even our industry where if you put something like pro agent, a virtual conversational agent in front of a consumer, people are like, okay, people are like, yeah, this is very similar to my experience talking to my phone talking to Siri or talking to like a Gemini and that has that overton window of what was, nah, it's not going to work. There's a little bit of pessimism to it has suddenly shifted in the last 18 months to be like, Hey, I have had a good experience talking to Gemini. had a good experience, let's say talking to Progressive and maybe I'll try also talking to this AI agent, even if it is like a collection or a law firm specific AI agent.

And we've seen that, like just like Matt said, we've seen that overton window of acceptability of what technology is acceptable shift radically. And maybe I'll touch upon two things. asked the consumer, like, how was your experience talking to it on a 10 point scale? And our average is like 9.7. So sure, once in a while, it is not perfect, but very off. Like most people are like, yeah, this is 10. I love it.

And so that has shifted. And the second thing is also it has shifted in surprising pockets. The fact that AI is more patient, the fact that AI can actually move at your pace, the fact that AI has no like the behavioral psychology of like shame and stuff like that goes away has meant that a lot of people that you would think are less receptive to technology are actually finding it helpful for them in ways that you did not imagine. So so those are like three four new things where consumer behavior shifts have have happened like fast they are accelerating and also in pockets. Actually, I did not expect.

Adam Parks (20:26)

Interesting. Now, as we think about all of the different ways in which we're starting to use artificial intelligence within our businesses, and if I'm not mistaken, because of your different disciplines, you're probably on different systems of record. And how are you orchestrating the communication so that you don't lose context between those channels? We've pushed for Omni or Opti channel, whatever. We've been pushing for that for some time, but if the right hand doesn't know what the left hand is doing, what kind of risk does that create and how does that hurt the ability to deploy this tech?

Matt Maloney (20:59)

I could take that because that was a really big decision and eye opening process as we were going through all of this and looking at it. Because ultimately, and you hit the nail on the head Adam, I mean if you're going to have these different channels of communication and you're going to now have, whether it's a human agent involved in having a conversation with a consumer, or you're going to have an artificial intelligence agent having a conversation with a consumer, that agent either way needs to know what's happened on that account in real time. Perhaps the consumer just came to your portal and we need to know that they just came to the portal, right? And maybe they checked out and made a payment, maybe they didn't check out and make a payment. A text message just got sent and so I believe that, and I think maybe going back for one second, I don't want the audience to be concerned that well if I don't have this perfect, don't want to move forward with it. It's about progress, not perfection.

Because everything is not going to be perfect as you go on this journey to transform digitally and then start to implement the use of artificial intelligence. So I always talk about that internally. We're not looking for perfection. We're looking for progress, guys. That being said, we determined that we needed APIs for everything. We didn't want to move and operate in a batch environment where we were batch processing SMS files. email files and that stuff was being loaded at the end of the day and then we needed things in real time. So that if our agent, our artificial intelligence agent or our human agent is on the phone in communication with the customer, they know that hey, at 12.05 p.m. today, an SMS message was sent to that customer. And that customer clicked on that SMS message and went to our portal but didn't engage. We wanted to know that and we wanted our in real time and we wanted our agents, excuse me, whether human or artificial intelligence to know that so that, you know, we could be not only from just a consumer communication so that we knew it in real time, but also from kind of a channel orchestration on the strategy side. We, like for example, we wanted to know that if Adam Parks did come and click through on a text link or an email link, come to our portal, but didn't check out or didn't make a payment, we wanted to know that so we could follow up within a certain matter of minutes or hours, whatever it might be, to be able to follow up with with Adam Parks and say, hey, Adam, we noticed that you went to the portal, but didn't didn't make payment or, how can we help you? And so it took us a while. That did not happen overnight. You know, we were batch initially, and then we moved to an API process. We had to bring all these systems together because the collection system that we're on is not it's great collection system, but it's not, you built necessarily on the latest technology where everything is there so you had to you had to work to build all and orchestrate all these different plugins to make it happen.

Adam Parks (24:00)

A coordination layer. Because if you're trying to communicate to all these different channels, regardless of where the data is, you're going to have to create some sort of collaboration. You're going to have to bring things together, which is, I think, the intelligence layer that we've talked about previously, and being able to pull those things into a single place in which it can be understood. Kyle, what's your experience been like?

Matt Maloney (24:01)

Yeah.

Kyle Cohen (24:21)

Yes, I would say I agree with Matt on a lot of what he said. It's getting to that point. And it sounds like Matt and his company made lot of progress on it. I'm sure it took a lot of time. And it was very difficult to get all those things to communicate. And it's going to be. And that's OK. I mean, that's something I'm working on. The idea is, and what I want people to know, is just it keep in mind where the world is going and where you want your business to look and make the decisions now today with that in mind. Like when you're looking at these new technologies or you're figuring out what to work on, it doesn't all have to be perfect. But make the decisions now that will push you in that direction and know what's out there. One of the biggest things I think people could do is, like if you haven't

Matt Maloney (25:02)

true.

Kyle Cohen (25:13)

like we, because of a lot of the legal implications, we've been a bit more hesitant for Voice AI. However, I've had several meetings with vendors and listened to several demos. Just take the meetings, know what's out there. You might be surprised. And another thing that AI has the ability to do and that additional communication has the ability to do is when, you know, this webinar is about an ecosystem. What an ecosystem can do is, you know,

Adam Parks (25:20)

Sure.

Kyle Cohen (25:41)

Did somebody log into your portal? What did they do on your portal? How long were they looking at it? Did they look at the options? Did they, you know, which option did they click and maybe not complete? And then that allows you to then go back and reach out through a text or an email or a follow up of just saying like hey is there anything we can help you with? Maybe we can do more on your account, or maybe have you consider this. And if you don't have that communication, then you're missing out on that opportunity.

And know, Matt was describing so many situations where it's just all these little things and opportunities can add up. So, I mean, we've definitely started our journey on ecosystem and we have certain things communicating live, certain things are still batched due to system constraints. But, you know, we're continuing to work in the direction of having everything work together in a cohesive system.

Shantanu Gangal (26:42)

If.

Adam Parks (26:42)

How did you, as a law firm, how did you prioritize the AI use cases within your firm? Because you've been more hesitant on voice AI, which makes a lot of sense, but what did you prioritize?

Kyle Cohen (26:56)

So we went pretty slowly. just for disclosure to everybody watching, we actually work with Prodigal. And it was a very long process of us signing that contract and where we are now. And that's because we wanted to be very careful for each step that we added.

Which is I would recommend to anybody so as first we wanted to just start presuit cases because pre suit cases is where the industry I think has we're basically a collection agency pre suit which is the threat of litigation. There's nothing that much more complicated and then okay now let's add in Some testing on litigation files now. Let's add in you know some testing on on more and more so And to be perfectly honest, I used to ignore prodigal's phone calls. So I don't think I've told Chanty that before. But you I kind of realized, hey, I just really need to see what's out there. And I'm glad I did because it's been really big benefit to the firm. And when I had that meeting, I know a lot of people get sounded by emails, phone calls. You might be surprised at what you see in a meeting. You might not be. I've wasted plenty of times in meetings that weren't the right fit. But I'm glad I still attend those because I didn't realize what they were doing and what they were working on. And I would say it's also just be prepared for any of these vendors you bring on that it's going to be a lot of working together to get this going. Because nobody's been doing this for 10 years, right? Maybe even working in AI for 10 years, but not what the technology looks like today. So I had to put a lot of work in with his team.

Making these guardrails getting these set up. mean we work together very closely So I would say make sure that you go with somebody who's willing to put in that time and effort to work with you really closely because the technology a lot of it's very brand new especially in this space where people have been figuring out the compliance of it and It's it's been worth it for me So It could do just little things that we couldn't do before. So as a wall firm getting into this, I would say just make sure you're putting in a lot of work and a lot of open communication with whoever you're using to make sure that there's no issues and they're following the guidelines and there's tons of things that we have and then, we actually need to change this. we need to change, like I'm in constant communication with this team of changing things and fixing things to make sure that the requirements that we have as a law firm and our clients' requirements of us and ethical, all these obligations we have are being met.

Matt Maloney (29:25)

Hey.

Kyle Cohen (29:43)

a lot of prefer it, right? We have one of the biggest complaints we used to get is when we would sue people, they would say, well, why don't you just, you know, give me a text or an email? Like I would have just resolved this earlier. And we have a lot of people that are actually very grateful that they do receive this text or email because they get to. They're going to not get sued. Nobody wants to get sued. So it's been a benefit. And if somebody doesn't like it, then they just tell us to stop. And then we just plug that into our system. So I'd say the big difference between a law firm and a non-law firm is as a law firm, we are the consequences. It's not like we reach out and you pay us or you don't. It's you pay us or we file a lawsuit, we enter a judgment, we seek to enforce that judgment. I would rather settle with people all day. And a lot of people prefer that. So I would say, again, the same words that people keep using is meeting people where they want has been a big benefit for us. And I thank them as well.

Shantanu Gangal (30:50)

building, if I may just build upon what Kyle said, a couple of very interesting things. One is there is a very clear realization that what we want to do, not just we as a vendor, but just like we as an industry, as a community, what to do is effectively overhauling maybe like 30 years of what worked. Like let's say from like even like 1995 to 2025 things worked in a certain way and obviously there were a lot of innovations on the margin but by and large everyone was doing internal emails and external calling and that was kind of the state of the world. Now call centers started looking different, you went from your basement to a cloud provider and so on so forth but by and large it is 30 years of systems that had not really changed drastically since 1995 and probably most people started using some of the earlier email plus phone systems. So that's what we're looking to change. As a result of this intelligence layer is brand new and it is has to catch up on like 30 years of best practices that pre-exist. So a lot of the close communication that Kyle mentioned absolutely like super important and it is like teaching a child. is teaching, it is being like a 30 year old system exists, but now you have a new child. So I think there is a lot of that tutoring parenting that needs to happen and that happens through some of the close communication and we are not shy in putting in the effort in order to do that and just like in any kind of parent-child relationship this itself probably has two sides to the same coin. One side is intelligence for all of us the humans who are actually coaching. And then the second is intelligence for AI itself. So there's intelligence that AI builds that is also consumed by AI. So it is like machine to machine AI. And we really like want to be very thoughtful about it because the matter of fact is a lot of what we do tomorrow as a group, as a industry as a civilization maybe even is going to be like machines talking to each other with next to nothing human in the loop. So we want to make sure that we have the right cartwheels. We feel comfortable with what the machines are saying to each other and we are able to kind of obviously pull the rip cord if it were to come to that but even when it's not we are able to kind of make sure that things are operating in a way that is obviously compliant regulatory but more important driving business outcomes. And so this intelligence layer that is effectively like overhauling like 30 years of best practices in the previous stack now have to be building that with close communication and even that itself has two sides to the same coin. The human to machine intelligence and then the machine to machine intelligence and being like super thoughtful about it actually unlocks crazy amount of like upside. It's just unbelievable how much upside you are able to unlock with that.

Adam Parks (33:54)

And you got to climb a mountain to try and orchestrate all of that data coming in from all of those different legacy system silos, bringing it together and being able to interpret it for next action because you're going to have to, it's not like it's a one time bring it over.

Shantanu Gangal (34:01)

for sure.

Adam Parks (34:09)

I think you have to build those orchestrations, which I'm sure is not an easy task, but well worth being able to integrate into multiple products, right? So a single point of entry for the data and kind of simplifies it in terms of its ability to orchestrate. But I'm curious because when I started asking the question to Kyle, I was thinking about like, so how are law firms actively prioritizing what they're attacking versus an agency versus a debt buyer, et cetera? And so I've started thinking about, as we went through the debt collection industry report this year, we saw that like debt buyers, for example, were more likely to be using AI technology for scoring type situations, whereas you had law firms that were using it more for back office. You've got the collection agencies that were very, very, very focused on communications, whether it be chat written communication or the chat AI bots. Have you been experiencing that, Shantanu, as you work across all these different disciplines of debt? Like, are people prioritizing the use cases based on their discipline?

Shantanu Gangal (35:17)

Yes. That is absolutely right. The way I think about it and probably is a function of what we see in the industry, different people have different kind of goals and the levers that move those goals are different. For a dead buyer, especially the passive dead buyers, they do not have a lot of team that is kind of captive to them typically at least.

So that's that. Their big lever is pricing. Like how do you price correctly? How do you segment what you're buying and how do you price accordingly so that you're buying stuff in a profitable manner? So that's level number one, pricing. And then level number two is time to liquidation. Obviously, till about 2021, interest rates were low. think no one really thought about time and interest kind of component of things as severely as people have been thinking in the last two, two and a half years.

If you can collect something in seven days instead of 21 days, moves the needle. If you can collect something as a dead buyer passively without actually ever having to place it in the network or ever even having to send like a demand letter that moves the needle massively. So that's on the law firm side. Very quickly on the agency side, typically agencies have larger head counts and that leads to that lever like optimizing your performance of your internal team, figuring out how do you drive your lot more self-service revenue is a big lever for agencies and law firms obviously have like very high costs and then the cost profile kind of changes depending on if it is pre-suit, if the suit has been filed or if something has you actually had a like if you're going to go and garnish something when you do it, what stage you operate. So there is different kind of costs there.

And depending on all three of them, I think the AI application is very different, right? If you are just, let's say a collection agency, you want to optimize your team's performance. You want to make sure that they are kind of being very productive to whatever extent any low-hanging fruit is being self-serve to the greatest extent possible. That's where you see a lot of like fully autonomous virtual agents making a very big difference. We see a lot of like digital texting, emailing, driving activity to the website and collecting dollars there because that is a very big lever for collection agencies. Debt buyers, not as much headcount. They're really focused on like segmentation, pricing. How do you buy stuff? How do you price? How do you segment your stuff and like which agency you match to their respective strengths when you place a portfolio. So that becomes a big use of AI. You also see some agencies obviously really focusing on the first seven days or every time they are pulling something from primaries and placing it to quads or secondaries and seconds quads. That interchange can lose like seven to 15 days and you really don't want to lose it. So I think you do a lot of smart things to minimize any time losses and money losses in that window. And then law firms obviously have different kinds of expenses, but the biggest expenses like filing a suit on something that you could get away without filing a suit or

Shantanu Gangal (38:27)

having to garnish something that you could actually have just use the fact that you have a judgment to drive in revenues. So optimizing for that using both obviously email and text to indicate clearly in a very compliant manner, but then try to level off next steps that a person needs to do to preempt that next big cost level kind of cost jump drives a lot of those things and those are things you need to be like super thoughtful about so that firstly it is compliant, but also because you want to get the timing right. If you send an email unnecessarily that's really got nothing to do with your of your cost profile, no one's going to act on it. If you send it too late, then you already incurred the cost. So you're kind of getting it just in time so that the person sees it, opens it, clicks it, goes to your website and resolves it before you incur like a $500 to $1500 cost is substantial. So as a result of it, think different people have different jobs they want AI to do. At the same time, everyone wants the next guy downstream from them to be aware of everything that has happened and kind of get that context as seamlessly, as quickly and as structured a manner as possible, which is again where kind of orchestrating across different jobs that AI does becomes very valuable. Like what you buy for What do you buy for cheap? want to work differently versus what you buy like kind of more expensive or like more valuable people. You want to work differently.

Adam Parks (39:59)

think that comes back to the notes. when I first heard it, the first time I had heard about your Pro Notes platform, I think it was for Matt, because I was writing a press release or something for Matt and we were talking our way through like what that tool set actually looked like.

But that was the first time I saw somebody starting to try and take the unstructured world and structure it so that it could be fed in the future. But I don't know that we would have the capabilities we have today if we didn't start, like Matt said, with those stepping stones of those small people. It's those little things that continue to build up and then we kind of find our way down the journey. But if we're not walking down the path, we're never going to see it.

Matt Maloney (40:37)

Yeah, yeah, that's why I was like I said before like we're looking for progress, not perfection, right? And if we take steps, we're gonna get there, as Kyle was saying earlier, just start to put yourself out there, look at things, explore what's out there, start to implement, it's not gonna be perfect right away, but, and I wanted to kind of, I broke off there for a second, we had like a power outage or something here in my building, so I got disconnected for about 60 seconds there, something like that, but we kind of look at, and this is a little bit to what Shantanu was talking about and maybe perhaps what your question and comment was out of a moment ago. We look at our operation similar to a restaurant. And so, and I've said this before, you have front of the store operations in a restaurant and you are front of the house as they like to call it and back of the house. Back of the house is going to be your kitchen where the recipes are made, the food is cooked, know, deliverability, all that stuff, right? And then front of the store is and the restaurant obviously is delivering the food and the touch points with the customer in our business that's collection, right? Collecting the account, engaging the customer. And I think we talked a little bit today about front of the house where artificial intelligence is impacting front of the house pieces of how we're engaging, how it's helping us make decisions and our conversational AI like the pro agent product that prodigal puts out. But in talking about scoring and segmentation, that's kind of back of the house, right? That's where the recipe is made. And everything that we look at today is kind of from a precision recovery perspective. Like I think it's a, you know, our businesses are evolving from a, you know, volume driven process into a precision driven one where like data analytics and all of our decisions and our strategy, our next best action is all driven by artificial intelligence, right? Machine learning and artificial intelligence are going to help us make those decisions. To your point, Adam, and you know, our agents haven't taken notes for years. I don't know how many years it's been, but it's for years. And so artificial intelligence sits over the call, listens to it. We had to build this process where it summarized and put the notes in like we wanted it to put in so that we could then and information out of those notes and then use them for our advantage to be able to build strategy, a precise strategy of what we wanted to do at the account level. so, within our company, we focused a lot the last 18 months-ish on kind of front of the house operations and how AI was engaging with the customer, making sure that it met all the compliance requirements but also just the fluidity of the conversation on the conversational AI side. We're kind of shifting our focus right now back to the back of the house and trying to figure out how we can have better just overall strategy orchestration and how AI is going to drive all of that because it certainly can do it a lot better than humans can do it because it can ingest a lot more information and process it a lot quicker.

Adam Parks (43:35)

That makes a lot of sense, you took the unstructured, you made it structured. There was so much value, there's so much gold in the collector notes that doesn't necessarily fall into a specific field in your system of record. And when you have that kind of intelligence that's available, but you're trying to find those ways to unlock it. And it was that first conversation that we had where I had that aha moment between, well,

Matt Maloney (43:39)

Yeah.

Adam Parks (44:00)

If we could structure the unstructured, or if we can at least start structuring into the future, maybe we can't go backwards for 20 years and do it, but maybe if we started today, what's that gonna look like in 10 years when we're looking back over our shoulders?

Shantanu Gangal (44:13)

And Matt, if I may add, the beauty of the operation you're running is now at this point, whether a note is taken by a person, like back to your restaurant analogy, whether a note is written by

Matt Maloney (44:13)

Yeah.

Shantanu Gangal (44:27)

someone serving or person serving on the front end or if it's a machine serving at the front end, the note looks exactly identical. I mean obviously it is marked by who took it but the note looks very identical. So when it comes to the back office, when you have to go and look at your segmental analysis, you actually have structured information that you can always separate by like a team versus machine but you kind of get a singular view and that just like unlocks a lot of analytics that was hard to do before. was like, everyone was like...

Matt Maloney (44:57)

That's right.

Shantanu Gangal (44:59)

everyone didn't do as much like analytics before Excel came along and like once Excel came along I think the cost of being smarter reduced everyone would be smarter and I think that is what AI is doing here and like having a very standardized way allows us to look at even sitting in the back office look at things so much differently one important part of the restaurant is the loading dock right you get you get deliveries and like optimized loading dock is a very big opportunity for a lot of restaurants Again, back office, like optimizing what you buy is probably like analog there.

Adam Parks (45:33)

But so the orchestration of that data and the ability to convert, let's call it those raw notes into usable data sets to build our actionable intelligence on top of, then comes the question, how do organizations...

Kyle Cohen (45:34)

Bye.

Adam Parks (45:47)

prepare themselves for that type of orchestration. from, know, from the podcast that we did a couple of months ago, it sounded like that intelligence layer can be laid over a variety of different data sets or silos in order to pull everything in. But it appears to me that absent of some sort of an orchestration or intelligence layer that actually operationalizing these great theories is challenging to say the least because we can't carry context.

Such a core problem, but that seems to be where that challenge lies. So guess, Matt or Kyle, from your perspective, how have you looked at it as you've started to go down your AI journey and select the pieces that are going together? Are you looking specifically at that cohesive layer for orchestration? Or how do you view it?

Kyle Cohen (46:42)

So something that was mentioned that as a law firm, we're very different in kind what we're looking at from an AI perspective. And a lot of it is just back office, operational work. Because a lot of what we're doing is we're just processing lawsuits, moving for judgments, processing garnishments. And A lot of those decisions made now by people are what are they doing? They're analyzing the data, right? That's the data inputs. They're taking it. They're analyzing it and doing an output, right? That's what AI does. That's what I do most of the day. I just see what's in front of me based on this. This is the decision I make or this is the action I take. And as a law firm, what might those look like? It could be redacting documents, organizing documents, preparing suits. Which suit do I prepare? Which template do I use? What should be on it? Is there one defendant? Is there two defendants? Are there multiple claims? Is this a loan? Is this a credit card? So it's all these decisions that are currently being made now. And then I like to think in my ideal world, how is this stuff going to be done?

And in my ideal world almost all that is being done by an AI who's seeing these documents, analyzing this, knowing the rules, making the correct decisions, right? And I think, you know, what could my firm look like in five years? And I think in five years it'll be an AI making almost all those decisions. Does it look like that today? No, it does not look like that today. Does it look better than it did a year ago? Yes, it looks a lot better than a year ago. Will it look better? a year from now than it does today? Yes. So what decisions can I be making to constantly improve, start adding on some of those capabilities and getting the tools in place? Like we recently just got a new copy machine. When I was meeting with the different potential copy machine vendors, I would ask questions along those lines. What kind of OCR capabilities does this have? Like how does it do this? Can it do this? And then now I bought a machine that those capabilities. Our old machine didn't have those quite as well. So now when I eventually go to have an AI be able to recognize these documents, redact it, or do whatever, I'm getting the tools in place. making the decision. I think that's the big part is you're not going to be perfect now.

Really imagine your dream scenario of what things could look like. I think all of that's gonna be possible sooner than we all think. And just what decisions can you be making today and every day to get you in that direction? I think, it sounds like Matt's way ahead in that journey, he's been doing that for six, seven years. However, maybe you haven't been doing that for six, seven years, but in six or seven years from now, the time's gonna pass anyways. So what decisions have you made to get in that direction.

Adam Parks (49:36)

What decisions are you making today that are going to prep you for the future? Go ahead, man.

Matt Maloney (49:36)

And I think it's going to go very, I think it's going to. Yeah, no, was just gonna say I was just gonna say that I think it's gonna it's going to get exponentially better in a much shorter period of time, right? So whether you're starting your journey today or you started it two years ago or whenever it was, I mean, it's going to get exponentially better very quickly. And that's why it just takes some action on our part to be able to just take that step forward, to be able to engage. There's a lot of great tools out there. There's a lot of great systems. There's this whole, we had this tug of war, like a buy versus build right and we did a little bit of both. built some and we bought some. because buying it from companies like Prodigal and other companies allows you to work with an expert partner and accelerate your push into that journey. But then ultimately because of all the uniqueness of your own operation like Kyle being a law firm and us being a first party service or a third party service or a debt buyer, we've kind of got and unique little aspects to our business, you have to customize some of these things, right? Because there's not really, and I wanted to say this to the audience, there's not really one size that fits all. Like there's not something out there that's right out of the box that you're gonna be able to open up that package, plug it in, and bam, you're there. It just doesn't work that way. It's gonna require your team to be involved over and continually be involved throughout the journey because there's different things that are nuances you within your own operation. You know, we're in the car rental business. We do all the accident claims for all the big car rental companies. That's a unique business. It's not like collecting on a credit card account. It's different. And so, but I do think to Adam's point, like, The recovery operations of tomorrow and maybe even today are going to be defined by this intelligence layer, not the intensity of how they do things with a resource, but this intelligence layer. That's going to drive it where AI determines the right action and the right channel and the right moment to do things. And AI will help process, in Kyle's example, and we have a similar example in our insurance subrogation where there's just a lot of documentation.

You have to be able to go through tens of thousands of pages. And it took an army of humans to do it before. We're not there today where all of it's automated. That's not a reality. But we will get there soon. And it's very exciting.

Adam Parks (51:59)

It's definitely an exciting time to be a debt collector. And into our final two minutes here, Shantanu, everyone else has had a chance to kind of rind out their thoughts. I thought I'd give you the same opportunity.

Shantanu Gangal (52:08)

I think again building upon what Matt and Kyle said. things are changing and the pace of change is also accelerating. So I think of the intelligence layer as driving some of these things and we are still in the very early stages of planting the seed of this. So once you have all the seeds in the right place, a lot of the acceleration is actually ahead of it, but you probably need to start doing some of it sooner than later. The other thing is again, the intelligence layer itself you need to think about it one from what is the intelligence that we as humans need from our operations that AI is going to provide and second is what intelligence does AI need from AI in order to build some of these things and accelerate its own development, accelerate the impact that it drives in our businesses.

And so that's that second layer, I think is not as deeply thought about is something we think about a lot is not as very widely spoken about yet, but will become super relevant. Like how does one AI regulate another AI? How does one AI prepare the work for the second AI to really like hit it out of the park? Those are the kinds of things that we will start hearing a lot more about going forward and something I'm excited for because again, back to the top, a lot of folks have experienced how AI can drive like convenient, compliant, more profitable growth.

And so to that extent, think that that is a very monotonic shift that we have seen at multiple points in the human civilization, industrial age being one, the printing press being another, the invention of fire kind of put us on a track to grow our society that we never look back. And once we learn to cook our meat and cook our food, we just became a very different society. And I do believe that this is the inflection point in similar league as those.

Adam Parks (54:11)

Well, this has been a insightful discussion today. I really do appreciate you coming on, sharing your insights with us and spending a little time with me. For those of you that are watching, if you have additional questions you'd like to ask Shantanu, Matt, Kyle, or myself, you can leave those in the comments on LinkedIn. The LinkedIn Live will be running the replay on YouTube next week. So we'll get that reshared so that more people can hear this really detailed discussion. I appreciate you guys joining and providing us with all of those different perspectives, because I think it really helps us to round out where this industry is headed and what challenges we're facing on our.

Shantanu Gangal (54:48)

Awesome.

Kyle Cohen (54:50)

Thanks Adam.

Shantanu Gangal (54:52)

Thanks for having us here. Thanks Matt, thanks Kyle.

Matt Maloney (54:52)

Thank you.

Adam Parks (54:52)

And thank you much appreciated and thank you everybody for watching today. We appreciate your time and attention. We'll see y'all again soon. Bye.

Matt Maloney (55:01)

Take care guys.