<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1308192452940245&amp;ev=PageView&amp;noscript=1">

Episode 16: TRaiCE’s Approach to Proactive Risk Management


   Kingsmen Beyond the Build YouTube Channel

Listen to the episode: 


Watch the episode:

About the Episode:

In this episode of the Beyond the Build podcast, Host Kevin Carney discusses with Sony Gabriel and Joe Kurian, cofounders of TRaiCE, about their AI application in risk management. The discussion explores TRaiCE’s innovative approach to leveraging unstructured data from diverse sources like news, social media, and customer reviews to predict risks in real-time, offering a unique additive tool for risk managers. Key insights from the development of TRaiCE, its integration within the banking industry, and its future potential in financial risk assessment are highlighted.   

Key Takeaways: 

  • TRaiCE utilizes AI to transform unstructured data into a structured risk assessment tool, enhancing the proactive management of risks. 
  • The platform integrates seamlessly with existing banking processes, offering real-time updates and predictive insights without replacing current systems. 
  • TRaiCE is gaining traction globally, with applications in multiple financial sectors and plans for further expansion and product development. 


Learn more about our guests:  

Sony Gabriel brings over 25 years of dynamic experience in the IT industry, spanning diverse roles such as Pre-sales, Business Development, Marketing, Consulting, and Program Management. His expertise lies in the Telecom and Financial Services domains. Sony's journey includes contributions to product startups like Metratech Corp. (now part of Ericsson) and giants like IBM Global Services. He holds a Bachelor’s in Computer Science and Engineering and an Executive General Management Program from IIM Bangalore. 

 Joe Kurian is a seasoned credit risk analytics professional with over 22 years of experience . He is one of the cofounders of TRaiCE Inc, a company that offers portfolio risk monitoring systems to business lenders using AI. Joe's extensive banking background includes working in  institutions such as GE Capital, HSBC, Wells Fargo, and MUFG where he was responsible for building analytical infrastructure within credit risk management functions, loan originations strategy, model risk management  and regulatory and internal audit preparedness. Joe holds a Masters degree in quantitative economics from IIT Bombay and M.S degree in computational and data science from Chapman University, California. 



Production Credits: 

Produced in partnership with Mistry Projects: https://mistryprojects.com/ 

Connect With Us on LinkedIn  


About the Host: Kevin Carney, Managing Partner of Kingsmen Software

Kevin Carney is a Managing Partner at Kingsmen Software. As Client Partner, Kevin assists clients in their transition from Sales to Delivery and then maintains a relationship to ensure successful completion. A Finance major by training, Kevin bridges the gap between business and technology, especially for Kingsmen’s banking and capital markets clients. Kevin has 30 years of experience in consulting to financial services institutions. 


About Kingsmen Software:

We are dedicated, experienced practitioners of software development. We grow software iteratively and adapt quickly to changing business objectives, so we deliver the right software at the right time. Our proven approach combines processes, tooling, automation, and frameworks to enable scalability, efficiency, and business agility. Meanwhile, our advisory & coaching services enable technology leaders, business partners and their teams to learn and adapt their way to sustainable, collaborative, and data-driven operating models.

Kevin Carney 00:04

Welcome to Kingsmen Software's Beyond the Build podcast. I am Kevin Carney, your host for today, one of the managing partners here at Kingsmen Software. Our guests today are Sony Gabriel and Joe Kurian, two of the cofounders of TRaiCE. Welcome guys.


Sony Gabriel 00:23

Thank you, Kevin.


Kevin Carney 00:24

So we're doing this remotely today, because we've got people in all different time zones. So Sony, you're in, you're in India, you're an ISP. Joe, you're in LA.


Joe Kurian 00:35

Yes. Okay.


Kevin Carney 00:37

So we're spanning What, like 12 hours of time zones. So thanks for joining us remotely. So I guess the question is, really, how did we enact if you're in India and urine, LA, how is the connection to Charlotte. So just give a little background, how we met, I saw a trace presented at Red tech labs venture 135. Last fall, and during the demo day, and it's it, it just struck a chord with me the offering that you guys have, because it's a novel way of using AI, at least in my opinion. And it's more additive to the credit risk management process. It's not trying to automate the existing process, it's additive to the process. So that kind of struck a chord with me. And, you know, it's stuck with me. But then coincidentally, one of your advisors Barry, Greenblatt approached me to, to join him in a panel conversation for the global association of risk professionals, Garp. And Joe, you are in the panel as well. And so like, Okay, this, like the universe is sending me a message I need to go talk to you guys understand more about your offering. And so why don't you guys explain a little bit about what trace does, and how you guys connected with Charlotte with ref tech labs, and just give them a brief brief overview.


Sony Gabriel 02:07

Great. Thanks for having us, Kevin. It's a pleasure to be on this podcast with you very excited. We actually got into the lab tech labs as part of class 18. It was actually an interesting time that we spend the Charlotte as part of the red tech labs program. Now what you do a trace, which got wrapped tech labs interested was how we monitor risks using unstructured data. If you know financial institutions have different risks that they have to manage credit risk, market risk, third party risk. Now the primary use when it comes to credit risk, at least they use structured financial data to monitor the health of the businesses. Now, you know that it comes with a lag. Now financial statements are usually annually or quarterly at best. They are the monthly Bureau reports and then the internal performance data. And high frequency signals you always get from all the digital footprint, which is aka unstructured data. Because it's non financial, it's in all kinds of format, text, video image, whatnot. Now, you can get leading indicators of risk from all of this unstructured data, which could be from news, social media platforms, review platforms, the L, the Better Business Bureau, and the like. And what we do at trace is leverage this unstructured data to find early indicators of press. And we go one step further, convert that to a score. So think of it like a FICO score, based on all the digital footprint. And this is high frequency so you can create a score every day. Right. And that's what traced this. And Joyce, the domain expert. I like joda time, and I just gave the high level overview of what we do. Yeah,


Joe Kurian 03:57

thanks. So thank you, Kevin, for having us here. Like you said in the beginning, I think that I did do nature, right, we are adding a new dimension into risk managers toolkit, you know, that is unstructured data, because that's also coming from my own experience, I've been in the business lending in last 13 years. So, the full spectrum like small business to consumer, sorry, the commercial lending to and their corporate lending space. So what we found is like Sony just explained, they primarily use the three pieces of data like bureau data, internal data, and the financial statements, but always this unstructured piece is missed out. So that is what we wanted to solve, you know, we, you know, go out to the internet and they extract every piece of information around the business, right? It could be discussions, it could be or news, or it could be reviews and other things and then use a large language model to convert that into a scorecard. So now, when there's a scorecard say scorecard like FICO, totally based on unstructured data, now, the risk managers know what to do with it right? Or you can blend it with other things. You can monitor it on a time series basis, say, okay, 20 point drop into scorecard and I need to pay attention to so we develop the whole application that fits into the workflow of a banker, you know, that's, that's so, you know, the specialized that, you know, that's the, the workflow integration is something that we offer in addition to a


Kevin Carney 05:43

Okay, so how do you go back to Charlotte? Like, like, what, what connected you to ref tech labs?


Sony Gabriel 05:52

That tech labs actually invited us to apply for the program? Did they, okay. And they were actually they have a team that scouts for interesting fintechs. We were invited to apply, we applied and then go through, we found the program very interesting. And very, it was very useful for us. That's where we found some of our advices we had a really good team of advisors who helped us build the right use cases got us the introductions. So the whole program was very useful for us, but that's how we got connected to Charlotte and connected to Barry, who is our advisor. Okay,


Kevin Carney 06:31

so yeah, they provide a little bit of ref tech labs provides little bit of investment they match up with with advisors. Those advisors hopefully get some network connections and introductions and industry. network effect. But, but but then you also have a little bit of a pedigree, right? You've got you've got no brand name behind you that you were accepted into that program. I think, I think Dan Roselli was telling me that the acceptance rate into ref tech labs is, is like, hard, it's harder to get into ref tech labs and is to get into Harvard. So that's a urine pretty, pretty good company, right?


Sony Gabriel 07:15

I think it's a six digit acceptance almost from the number of applicants. So how many make it to the program? Yeah,


Joe Kurian 07:23

I still remember. Dan was very much involved in the selection process in early on. And I think we stood out because of the FinTech sandbox, we know we were part we were admitted to FinTech sandbox, it's another accelerator program that they provide us or connect with a lot of data providers. So we could, you know, get data from those guys and try to develop new solutions. So I think admission to that was a you know, kind of a standing out feature for us. So rev tech labs, kind of, you know, invited us to apply.


Kevin Carney 08:03

Okay, so keep building on those. Right, you get those? Yeah, you get those items on your company resume. And


Joe Kurian 08:09

we were also looking for bangers as the primer writers solution is for Bangor, so Charlotte is bagging hub. So all that, you know, so for example, I would think you know, a lot of there are a lot of other accelerator programs, but rev tech lab was one of the no real good ones for FinTech, who is in the banking area, right. So you know, because of the access to the market.


Kevin Carney 08:31

Right, right. Okay. So how did you get to where you are today? Like, like, there's a, how did you get the band together? Right. I think there was a, your, your your brother, you went to school with your brother and you guys had a prior company and your advisors like give give the whole origin story? Sure.


Sony Gabriel 08:54

So do I know each other for almost 20 years, his brother and my brother and Joe went to college together. And other co founder Geeta and I, we did our undergrad together. So I know Geeta for the last 30 years. So it's a close knit founding team in the sense we know each other from the past. We also ran a company before this, where Joe was actually involved as an advisor. So Keith and I were the founders Joe was advisor. So kind of like the same team, which worked before as well. So in 2016, we founded Minerva software, where we primarily did product engineering, we will products for global clients. And the key differentiation was that we build products that leverage data. So initially, we build the web applications to collect all the data. Then we build AI ml models on top of it, and that was our gig. Now, interestingly, there was one client that we worked with a very large marketing firm out of Canada and for then we, we did a lot of work around sentiment analysis, natural language processing, which today is more popular as large language models and generic, but we've been playing with this since 2017 2018 timeframe, we built the entire product. And they went on to base a truckload of money with the product. And but gave us a lot of expertise in this place. And that's when Joe was like, there's a problem statement in my line of work where what you're doing with unstructured data, combined with, you know, knowing the business can help us build a product. And that's how we form tricks. Right?


Kevin Carney 10:41

Yeah. Oh, gotcha.


Joe Kurian 10:44

Just to add a little details there, I think I was there advised of that time. And it was also the COVID time COVID Just hit. And then I had this idea, and I was watching them closely. So they were doing a lot of good work in the area of unstructured data. You know, they started with IBM Watson, for example, right? It was like, you know, and then then switch to Bert, when, you know, it's a Bert was released in 2018. But 2020, these guys were kind of 2090 they were using it. Now, then I realized, Oh, this is this is great. So then I present this idea, then. So I think the team was way ahead. So for example, we launched the product, you know, almost 18 months before Chad GPD. And when we launched, nobody would know, you know, what is LMS. And we had to explain the word slurs, language models and things like that. So for a general audience, right? I mean, so that so way ahead of, you know, where everybody was. So that's, that's another thing. From outside of standpoint, I realized that, again, this team is great. And building building product, and especially in the very innovative product. So


Kevin Carney 11:59

that's a big differentiator, right, because a lot every company I talked to right now, they're all using AI and Kingsmen, as well. But the difference here is that you guys have been using it well, before it became popular. And I think you also use it because there's a distinction here between Gen AI and machine learning, right? I mean, you guys are using it on the machine learning side, which is a lot more complicated, and less accessible than it is on the Gen AI, right? It's easy to go to check up and create a term paper or go to mid journey and create an image. But the machine learning you guys are doing is Is it a much more complicated level?


Joe Kurian 12:43

Correct. And I think the common thing is the large language model, right, both of those, so we use a large language model to create an application. So Janay applications, right, like chat. GPT is another application, which is more elaborate, you know, a lot, a lot more elaborate. But this is more like a custom, specialized software meant for certain use cases. So that's the difference. common thing is large language model, right? That's


Kevin Carney 13:10

right. Yeah. I like the you bring together you know, Sony, you and get the head a, a capability that you're using for your clients and something that one client in Canada specifically, but I'm sure other clients as well. And then Joe, you came at it from a business standpoint and said, I need something I could use that capability over here in this particular business case. So pulling those two things together, you know, it's not I have a capability and I'm looking for a reason to use it. You know, you're you're actually using a direct scenario that that Joe coming up through the ranks of, of business lending needed, right, because like you mentioned in beginning Sony, the financials are looking off the back of the boat, you can you can see where you've been. This is where we were four months ago, because by the time q1 financials are, are in the books in the end of March, and then they're finally audited at the end of April. And then they get to the bank and that taking a couple of weeks to do the statement spreading, like anything that you could surmise from the information already happened. And it's not actionable. And so I love the fact that you said the high frequency that you can be proactive in either turning down a credit line or increasing interest rate or asking for collateral or invoking some of the covenants like there are things you can do. Yeah, I just I think it's great.


Sony Gabriel 14:45

And Joe have this number ones he said 95% of your risk comes from 5% of your portfolio, like now you understand that 5% of your risky portfolio by the time you do your quarterly analysis and studying and all of it, that picture would have entirely changed. The five persons that you that you actually had with those financial statements is no longer the five person because things can change daily can change as well, like, so that's where the value comes from. Right.


Kevin Carney 15:17

So just just finishing up the the origin story. You guys have got some pretty good degrees college degrees here, right? I think Sony, you went to IBM.


Sony Gabriel 15:31

Enjoy it. Yes. Can you can you guys


Kevin Carney 15:34

explain that a little bit? Because those are like the ivy League's of of India, right?


Sony Gabriel 15:41

They are. So I did my undergrad in computer science that were different. I did an undergraduate because we didn't want to the universities here in India. And then I went on to build my career. And then midway through my career, I did an executor management program at the Indian School of Management, Bangalore. That was in 2008, nine timeframe. Good. I went on to do her master's in computer science from University of Illinois at Chicago. That was 9798. So yeah, he's a master's in computer science. I did my management degree. And yeah, we'll talk about what has this spectrum.


Joe Kurian 16:23

Yeah, I had a conventional economics and statistics training, and I joined IIT in the Ph. D. program, I was admitted to, but I did not finish my PhD. I just did two years and graduate with an MPhil. But that is where I was introduced to something called SAS, I don't know you in the sense that critical analysis system. So I got really excited with all the statistical modeling, all that part. And so I picked up a lot of skills there. And that kind of changed my career from a conventional economics guy to analytics person. That's, that's where the change happened. So yeah, they going to that Institute helped quite a bit because of the infrastructure and other things. So in other words, I wouldn't have seen it. Maybe I wouldn't have ended up in GE or anything like that.


Kevin Carney 17:16

Yeah, you work for some great big companies, right, like HSBC, and Wells Fargo and GE Capital. Yeah, correct. Yeah. I guess where I'm going is that you two guys have and Keith as well have a really impressive background. You know, this is not just some guy with an idea, right? You guys have the educational background, you've got the business experience, either whether it's you're building a software firm with Minerva, or working in the lending space with some of these big financial services institutions. While this trace may be a new company, to a lot of people, it's it's it didn't happen overnight. There's a lot of hard work that got you to where you are, to where you get the right people together, you get the right capability, the right product, and it's all kind of coalescing into something that is starting to to really take off. So that's pretty neat. Right? So give me some stats here, like how many people how many clients capital raise, think things on like things that you're interested in sharing.


Sony Gabriel 18:23

So we are a rather small team. As a startup, we have a 10 member team out of India, which is where the product team is, and I'm based here in India with a team. That's apart from the three member founding team. So all in all around 30 member team. We have five clients and several paid pilots at this point. We have clients across the US, Germany and India. Three of the clients in the US one in India, and one in Germany. We have a payments provider in Germany that we work with for margin monitoring. In the US, it's community banks and credit unions. There's also a small business lender and primary use case is around portfolio monitoring, prospecting, and a lot of third party monitoring. So mentor monitoring dealer monitoring, those kinds of use cases, and India we work with, like the Vietnam EQL and the payment settlement infrastructure in India. So we work with them to monitor the participating banks on the fast payments network. So it's a case of a counterparty risk monitoring, kind of use case there. So those are at a high level, we have raised around close to half a million and mostly angel funds and then from the red tech labs. So that's where we are early days of revenue. So but we are seeing some interesting pilots. So that's basically where we are now.


Kevin Carney 19:55

Right? That's pretty neat that you are so global ALL RIGHT WITH US and Germany and India. And isn't there a potential for Canada as well? Finally going on in Canada,


Sony Gabriel 20:08

we are actually working with a large Canadian bank. Okay. So, but it's mostly for the US operations.


Kevin Carney 20:15

Oh, I see. I see. Okay. Okay. So just as a sidebar, how do you deal with sentiment analysis? In in those different geographies? Right. So you've got language difference, but then you also just have a sentiment difference, right? I mean, that's, are you retraining a model? How does that work?


Sony Gabriel 20:37

Joe can take?


Joe Kurian 20:39

No, not really, I think, primarily English. It's not just for the regional differences. It's English as like, as a I'm sure the training is done for us for, for example, the UK versions or US versions, but But otherwise, my understanding is that it doesn't make a huge difference for regions for different things. Okay.


Sony Gabriel 21:08

So for languages, Kevin natively process most of the major languages, so with the German provided that we are working, they have the primary businesses out of Austria, but they have merchants across Germany, the Nordic Region, and most of Europe, so Denmark, Finland, Netherlands, and all of it. So we do see our news and reviews and all of these languages. And we process those natively, so. And so since, like I said, most of the major languages are natively supported. Now, for there are a host of languages, which probably are not natively supported, but then we can always do a translate and then do a sentiment analysis. Okay.


Kevin Carney 21:54

That's impressive. That's nice. Okay, let's pause for just a second. Let's do a bit of an icebreaker. It's always weird, the icebreaker in the middle, but it's a fun part. Okay, so the news and AI that's coming out all the time is always so fantastical. It's like they I see headlines every day, whether it's from some of the sites that I monitor, or some emails that come in. And I look at some news articles, these headlines, and I'm like, That can't be real like that. That was totally manufactured headline that is not right. And look into it. Like, oh, my gosh, it is real. So So I have a few headlines here that there's six of them. Three of them are real. Three of them are fake. And you have to figure out which ones are real versus fake, of course, what did I use to go create the fake ones, but you know, chat GPT because that worked pretty well. Alright, so here we go. I'll start off with one. Number one, cognition has released AI software engineering agent, Devon, that is capable of autonomously delivering software.


Joe Kurian 23:06

That's true story.


Kevin Carney 23:07

True. Yeah. Do you think Sony?


Sony Gabriel 23:13

If your says to


Kevin Carney 23:15

that, do you know? Do you know what's true? Or you think it's true? Joe, do you know what's true? Or do you think it's true?


Joe Kurian 23:21

I think I read this, Devin.


Kevin Carney 23:26

I know it's true. I shouldn't have said Devin. Yeah, that's true. So we've been looking at Kingsman and it is pretty. It's not available to us yet. But some of the demos that they're showing are pretty impressive. Where? Yeah, it's a software engineer in a box that just it'll take a requirement and it'll start building software and it will build tests and make sure those tests work and it'll iterate on itself and explain to you what it's doing and why. It's, it's, it's really impressive, and I don't know what it means for Kingsmen software over the next year. Like that's my challenge is like, up there, maybe you know, it needs us anymore. Okay, all right. So you got that one, right. All right. You can score I don't know. Okay. All right. Broad, sorry, Freud, AI, Freud AI, I guess it's like Sigmund Freud. Freud AI detects unknown human emotions, expanding psychological models. So I guess his AI was was finding patterns in human like you met. You mentioned sentiment analysis. So Freud AI detects unknown human emotions, expanding psychological models. Real or fake?


Sony Gabriel 24:47

I think that's fake.


Joe Kurian 24:51

I think it's real again. Yeah.


Kevin Carney 24:55

That was fake. It was fake. Whether I feel better or not. Okay, AI machines developed their own language, excluding excluding humans. So the AI machines have built their own language and they're starting to exclude humans from their conversations.


Sony Gabriel 25:19

Thanks. But I'm still not sure with no. Right. And, and that's when people start thinking about all this Terminator movies and other things that you let it loose. And then before you know you're out of the picture,


Kevin Carney 25:35

so I just I look at that and go, that is very feasible. What do you think? You got very real or fake? I think. Yeah. Sorry. What was that? You said? I think I think it's fake, fake. It is fake. It is fake. But it's not


Sony Gabriel 26:00

unrelated. Right. But I read that, you know, here's the problem with all hallucinations that we talk about, right? You see it written in so many places, you come across it? And then it's very difficult to distinguish whether it's real or fake. Yeah. That that that it's it's very difficult now it's, it's just so difficult.


Kevin Carney 26:27

I wouldn't I wouldn't put it past a I start coming up with his own language. I we're using AI very generically here, right. Like, like AI is not a thing a as capability. It's not like there's one AI thing that's doing stuff, but I can certainly see. A I start to like, like, the English language has to be fairly inefficient at some level, like, clearly that's why it can be great to talk in ones and zeros. But there's probably some other way to convey. Thought, then English language, right? It's not, it's not an optimized language. Anyway, all right, number four viggle. Di G G, l e. Allows for 3d video, face and body swapping. 3d video Face and Body swapping.


Sony Gabriel 27:39

I believe that that's real. Job, but I realized, yeah, wouldn't it but it sounds weird.


Joe Kurian 27:49

Yeah, I I'm not sure. I think Yeah. Yeah.


Kevin Carney 27:55

So I just saw this last night. There's memes going around social media. So there is a rapper, little Yachty. That kind of comes from backstage on the stage to some big huge crowd of people. Like when it's one of those stages that kind of juts out into in the audience. And there's so this is real CT. Like, this is a real headline. So at the top, it shows little Yachty coming from backstage going out into the, onto the stage, and all these fans and he's like doing his dance and whatnot. But people have been faced swapping that with like, Elon Musk, and an astronaut uniform. Or the Joker from from Batman movies. So like, like, now everyone's just based in body swapping into anything. And so I can only mean if only I can only imagine what this might mean for our election cycle or presidential election cycle coming up soon. You know, someone's going to put Joe Biden or Donald Trump on this little Yachty meme and we'll never know what's real anymore. It's It's It's crazy. Okay, number five, AI helps archaeologist decode ancient scripts


Joe Kurian 29:28

Yeah, this is a deal.


Sony Gabriel 29:29

That I think it's possible.


Kevin Carney 29:31

Yeah, that's real. Yeah, I saw that too. It's like a Rosetta Stone right? It's it's sort of like ancient Sanskrit or something from like the what's the the Babylonians or something like that and they're able to figure out what the actual characters me mean, right? Because these are all picked geography. Yeah, I don't know how they would do that. Like, I can't even imagine how that would actually work. But it does. But then again, I don't know how people would do it normally. Like, how would you do that manually? Because they figured out all sorts of like, I hear about the Rosetta Stone. I don't exactly know how the Rosetta Stone actually worked. But all right, last one. Ai wins Nobel Prize in literature for the first fully AI written novel.


Sony Gabriel 30:39

Let's be reading a novel Frys that sounds fake. Okay,


Kevin Carney 30:47

it is fake. But you know what's gonna happen, right? It's gonna happen, which just have to be we have to be ready for it. I think that's why all the Hollywood was striking over the summer. Right is because or, or over the winter is because they were worried about AI taking over their jobs. novels, too. Yeah. Okay. And I think


Joe Kurian 31:13

the if Nobel committee is kind of a drone, they replace all the politics out of it and put a machine learning algorithm to select the novel, then probably, we will have a winner from an AI.


Kevin Carney 31:27

Yeah, yeah, sure. They'll just like humans would vote for who they know, the AI will vote for what they know. Right. You know, one of the other suggestions that I got about a fake article was a fully AI jury convicted a defendant, like put them in jail. And you're like, that's scary. But then I thought, well, they would be the most impartial jurists? Yes. Yes.


Joe Kurian 31:56

I read. I read something similar to that. I think it was tested on parallely. They run the algorithm based on all the rules and then see how the judge come up with and see the differences since Yeah.


Sony Gabriel 32:09

The only caveat there is Kevin, it depends on how it's trained. Sure, if it's an impartial training, like you actually said, everything that was ever meant. All the cases ever done and the verdicts that came out, then they will probably be impartial to an extent. But if it were selectively trained, then


Kevin Carney 32:36

what if? What if you put all the existing cases and verdicts? Um, those are probably biased themselves, right. Like that's


Sony Gabriel 32:47

true. But you probably wouldn't be putting all the rules in all the legal framework as well. Right? Oh, I see. Yeah. Make that all up. It's it's there. And it could make more impartial, unbiased decision, then. Yeah.


Kevin Carney 33:04

Maybe each. Each jury should have an AI 13th. jurist. Right. If there's normally 12 Jurors give a 13th AI and see what they come up with. Yes,


Joe Kurian 33:14



Sony Gabriel 33:15

That'd be that'd be interesting.


Kevin Carney 33:18

Okay, let's get back to a trace here. We sidebar enough. So the the unstructured data and the sentiment analysis, what are you mentioned, you have a score, what's your what's your branded score name?


Sony Gabriel 33:34

Called the business sentiment index. We call it the BSI. Okay.


Kevin Carney 33:39

You need some fancy naming around that is some sort of trace TM kind of thing, right?


Sony Gabriel 33:48

To say,


Kevin Carney 33:49

but you something you mentioned earlier is that you're able to show the traceability, right? It's not just that the score went up or went down over a day or a week or a month, you get into the description as to why right or at least pointing towards Why would have gone down by the how do you do that? What does that look like?


Joe Kurian 34:10

Yep, I think first thing we do is, you know, get go to the sources and extract every piece of information that is relevant for the business, and then convert into a scorecard. And when we do that, we we do a retrospective scoring, right, like five years. So we create a score card for five years. So you have like a score for a month, every month you have a score. And now when so then we have like all these underlying articles, let's say if it's like a news article, or a review, or a discussion, anything like that, so we also have a link to that particular one. So if somebody wants to see say for example, think of it as like a scorecard. The score dropping from say nine 100 to 800 writes 100 point drop. Now you are curious to know, so automatically an alert will be sent to the portfolio manager. So then you are curious to know what what actually caused this drop. So you come to the platform, click on that month, and you can actually see what all our articles contributed to it. And also there are different weightages like, you know, for news from Wall Street Journal gets maybe 90% weightage, the review from Yelp or Better Business Bureau, there might be only 10% VT, so you can see all that in the platform. And then it's a full transparency, right? I mean, because bankers are very touchy about, you know, what, what is going into the algorithm? So, we provide that observability to end their pieces of information going into the scoring. And that so the answer, so that is very interesting to, you know, compliance. And also regulators as well, you know, some of the, one of one of the banks using us, you know, regulators gave a very good feedback, because it's, it's traceable to the last, you know, the most granular level like, which is an article or a Yelp review, you can go read them. So, it decide, you also check that it is relevant or not, and rescore it all that you can do in the plan. The


Kevin Carney 36:18

regulator's must love this, right, they must, they must really appreciate that there's something else going on, beyond just a normal conventional credit risk management portfolio management.


Joe Kurian 36:28

Yeah, in fact, you had meetings with the regulators, you know, so, you know, we reached out to them, and, you know, he wanted to see the demo. So these are the Yeah, this, they can see a lot of application, you know, like TPR Mizmor, the third party risk management, vendor monitoring is very, very hard these days, right? Every day, we know because of the bank, FinTech relationships and everything. So it's so useful because it's organizes the entire data around the business in a timeline. And then you can drill down and see the details. So that is very, very useful for the


Kevin Carney 37:05

Sony Do you want to add in there.


Sony Gabriel 37:08

So when you look at how bankers are using it, a portfolio manager or someone in the risk team, they they start the day with an alert, if there is one, imagine there is a news which had some red flags and you know, some fraud or something being reported that the team would get an alert saying, the business sentiment index dropped by a certain points and this red flag was noticed, which has how they said they they then they know they have to drill deeper to find out what happened. They can go into the application, they can see this the specific news, the link to the news, they can read the news, if there is more than one, they can see all of that. So we provide that entire traceability. We provide automated alerts to let them know when they start their date as soon as they are done. And that's how people are using it today. Right?


Kevin Carney 38:03

Yeah, it's just it's another radar, I guess, of knowing if you look deeper, right? So if you do see a red flag, like okay, well, someone left a bad Yelp review, not that big of a deal. It happens, you know, move on, but Wall Street Journal had an expose a on on this company, and it looks really bad. You need to pick up the phone and call that client and ask them what's going on. So it gives them it gives the bankers much more line of sight into what is going on with all their clients without having to physically follow them, physically follow them, but like, spend all their all their time and energy actively following them.


Sony Gabriel 38:48

And we have also said, it doesn't come into the mainstream news in the beginning, right. So we have seen a case study where there were a lot of complaints from users on Yelp and Better Business Bureau. And all that was happening for a month, then it became too loud that news agencies started picking it up. But when you're using trees, you're getting those signals early on because you're seeing all these customer reviews, which are pretty bad. And you know, as a relationship manager or someone in the recipe, what you would do is you would want to know why so many customer complaints are coming because news will not happen with the first customer review, it's probably going to take months for it to reach a stage where it's too large for a news agency to ignore.


Kevin Carney 39:39

Right maybe that should be another client base for you is all the news agencies to to bubble up information for them so they can write an article about that. We


Sony Gabriel 39:51

haven't thought about that but that's


Kevin Carney 39:54

alright, well that that takes off. I want to commission and absolutely So with your your typical client base paying community banks and in the US and others, but but you're when you're walking into a sales situation, what I like about your solution is that you're not trying to replace some other solution. Right? You're not trying to be better at statement spreading, you're not trying to be better at underwriting. Right, you weren't you are trying to add something to their portfolio management process. Does that make that easier? Because there might be some integrations there. But no one has to do a business case to swap out the existing platform. There's no incumbent, if you will, does that make life easier? Yes.


Joe Kurian 40:47

I think they instantly recognize this as an added dimension, you know, into the toolkit. It's not really replacing anything, right? I mean, because they, they still want to use other pieces of information, and this will be an added. So it's kind of, and also, they're doing something similar, like their underwriters are now going into five different or 10 different sites and trying to consolidate or Google search and other things, which is like pretty, pretty difficult process of identification. And now if you have like a 10,000 businesses in your portfolio, it's impossible to do it right. So I think they see the value of cutting down at least like 90% of the efforts around of color, you know, kind of going ahead and gathering information that is certainly there. And then more proactive, right? Where we are positioning this is the proactive account monitoring space, which comes before the delinquency right now, most of this community banks that we go to, they have a strong relationship banking, now, if they know in advance, something is going to happen. There are many mediations they can do, like, you know, say they can, you know, if it's like unsecured position, they can make it a secured position, like I get a piece of collateral or something, right. And then worst case is they can cut the line save, like $100,000 line of credit, and they can cut the line and say no more drawing from the line. That is another mediation strategy. Right. So in between that there are many things they do through relationships. So that is where we are positioning designs is it is really resonating with what they do on a daily basis.


Kevin Carney 42:25

Is that your target market, like small business banking, or you get into wholesale to get into commercial lending? Like what what's your love could be all levels? Yeah.


Joe Kurian 42:36

I mean, I think that's because they want to track and monitor a business we are getting, we found that in a day of corporate lending level of commercial lending level, we get a lot more data, compared to small business level, but then small business, they have this advantage of say, for example, a customer facing business like a restaurant, or you know, those kinds of dental places or professional services, you know, they they have a lot more review data, right, Yelp reviews, Google reviews, and so it's the blood, I think, you know, people are finding, you know, a lot of use cases in the lower end of the spectrum to like small business bragging as well. It's probably


Kevin Carney 43:17

harder to find that data to write, like, if your company is Boeing, you can easily see in the news that Boeing is struggling right now, but if you're, if you're like, a dental practice, like you said, yeah, it's harder to find those things than information. Yeah.


Sony Gabriel 43:34

Sorry, Tony, you know,


Joe Kurian 43:36

I will just going to finish up saying, you know, there was a McKinsey article recently about how this whole thing works, this works like a, this is like a binary switch, what we are providing is which is like an alert, you know, you you have like a sometimes there won't be anything on this business for a long time. But suddenly something happens, right? So then that that triggers the review mechanisms and everything we just worked at, you know, just keep monitoring all the businesses in your portfolio work that because sometimes this pipe will happen and and it's very, very useful for the because it you know, when you have very little information, but from even from a theoretical standards, recall information value is very high for that kind of small information, right. So that is what happening in monitoring like this. So God, sorry. Yeah,


Sony Gabriel 44:30

I was just adding equipment. That was a great point. Thanks for bringing that many. We are talking to a large bank and they have corporate as well as small business now, most of the large banks will have Bloomberg and other licenses, and Bloomberg typically cover all the large corporates now even the midsize ones. But when it comes to small businesses, the you don't have a lot of data points that you have today. Bank, the portfolio team still has to go to the internet and see if something What's happening and usually, we have 1000s, like Joe said, You cannot monitor, we cannot manually be tracking all this 1000s of businesses, right. That's where we come in, because we have automated the whole process. If there is something, we will pick it up, we will score it, we will send them alerts. So as long as they know that, you know, there is an auto, think of it, like your automated assistant, or an analyst sitting there and saying, Yeah, everything, everything out there, bringing it nicely on a timeline with the score and the explainable P for you to come to you. Yeah,


Kevin Carney 45:36

got me thinking that. One of my clients is it does does investment banking research, you know, sell side research. And they do have an analyst. They're not working with big companies like Boeing, they're working with smaller companies that you recently went IPO. And there's 10s of 1000s of those companies. And yeah, they have an analyst that's, that's looking at all that information right now. So I bet they could benefit from that. So you can focus on the more important things,


Sony Gabriel 46:09

right. So it's like, we're not trying to replace analysts, but we can make them far more productive. Help them focus on the right things. And even if they were to increase their portfolio, they don't need to increase the team. Now with the existing team, they can increase the size of their portfolio, they can go after more businesses.


Kevin Carney 46:31

Joe, you mentioned or alluded to the third party risk management. So if we move beyond the credit risk management or portfolio management, how is how can trace be used for third party risk management and vendors?


Joe Kurian 46:44

Yeah, this gave us a warning in our tech labs, you know, we, we had Theresa Theresa Burton. So she was our advisor. So it came from her and, you know, she said, You know, I am filling out all this, you know, checklists for vendor monitoring stuff. And, you know, please could add a lot of fill in a lot of pieces there. So that's, that's one thing. And then second second eye opener was when we installed the product in one of the banks here in California. So they said, Okay, monitor this, but also monitor all of our vendors, you know, like, so they gave us a list of vendors and you know, other other parties, they deal with them in the, you know, apartment associations and other things. So any anything that they sponsor events and everything, so they wanted to add the full list. So that's when we really realized, right, I mean, there's a need for this, because ultimately, we are monitoring businesses. So they wanted to see, are they doing okay, in the market? Is there any wrong news or bad media or anything on them? So it fits very well with the TPR. Um, as I also we kind of develop that a little bit with Teresa. So that's a, you know, we started positioning this for TPR, as well.


Kevin Carney 48:01

Yeah, that doesn't make a ton of sense. And we have things when we fill out those, those forms all the time, but it's a one and done kind of thing, right? Maybe it's once a year, but no one's paying attention. And if you're a, if you're a company that has 10,000 vendors and one of those vendors does something inappropriate or bad or something you don't want aligned to your branding? You want to know quickly that that's happening. Yeah, I could see that. Okay. What about you mentioned earlier, like dealers helped me understand that.


Joe Kurian 48:38

Yeah, that's another bank drone use case, you know, we are discovering this, I mean, same thing is used by because bankers are very creative, right, when they see some tool, they know how to use it for their or so, the dealer monitoring came as just like that, you know, we were talking to a bank and they had a large $1 billion solar lending portfolio. So, that is actually consumer really speaking, right? I mean, because you're lending to the consumers, but how consumers get this loan is through a dealer, you know, they go to the dealer for solar installation, and then the dealer will facilitate all that loan approval process through the bank, but ultimately, bank is lending to the consumer, but if the dealer is doing something, you know, say botched installation or you know, not providing the right service, the consumer will go and delinquent on the bank loan, you know, because they are not very clear like so, that is the indirect lending use case and which is like a dealer monitoring is very important for the bank because some dealers they found say things like pink energy, right, I mean, that is one that really went bankrupt, and dealers like that is causing disproportionate amount of losses to the bank portfolio. So they quickly understand that so now, if they know in advance those guys are getting into Trouble through Yelp reviews or Better Business Bureau reviews, they can scale down, right? I mean, and they can go slow on lending or even stop accepting loans through the dealer. So that's the use case. And if they go bankrupt, that's also a big loss to the bank. Because, you know, a lot of inbetween deals and you know, money's in between, so all that can be at risk. So that's another use case, we found very interesting to solve our industry specifically. And then it is true for even auto dealers, right when your auto dealers, banks watch monitor auto dealers. So that's, that's one thing trending pretty high these days. Yeah.


Kevin Carney 50:41

Get anywhere, you might have some sort of an installation that could go poorly. I don't know, like medical practices, right, like cosmetic medical services. And


Joe Kurian 50:50

yeah, yeah, that's the big one. Yeah. Okay. Interesting.


Kevin Carney 50:56

So how can I go get my Kingsman score, like is that also a market is like, because, as an example, I can go to Experian and see my personal credit score. And I can monitor why it's that way, like there was a delinquency many years ago or extend about like, I know how to improve it right, like Credit Karma will teach you how to improve it, is that the thought you guys have are.


Sony Gabriel 51:21

So there was a marketing that we had at some point where we said, Come and score your own business or bring five businesses you're interested in. Now, the thing is, you need some capital behind you to run, because what if 1000s of companies and you know, want to monitor this code, right? So at this point, the capability is there within the platform, but the it's more, it's a resources that puts a constraint for that service. Okay.


Kevin Carney 51:53

But it's doable.


Sony Gabriel 51:57

I mean, think about it. Like for personal investment purposes, when you are in the stock market, you want to track a few businesses that you're invested in, or you're interested in investing in, you can monitor all right,


Kevin Carney 52:08

yeah, sure. It's personal investor already wants to know,


Joe Kurian 52:11

yeah, sorry. If Kingsmen wants to know, the score, you know, they can always I mean, you know, we can always set it up, you know, without only thing is, you cannot just go there and see later, you know, if you will have to set it up right now.


Kevin Carney 52:25

Sure. Yeah. Yeah. I mean, sure, as things start to snowball, like, as more banks start to use it more businesses that are going to care about what their digital footprint is. And so they'll want to know what there's what theirs is, and how they can improve it. Yeah, right. Now, it's all very ad hoc. I mean, it's a funny story from a Kingsmen. standpoint. So there's, there's glass door, right, you know, the, where you can bring up your right your employee employees can rate employers. And we received a mediocre review. And like when we wanted to respond to it and see what, you know, what someone's concern was, and maybe we can learn something from it. And it said, yeah, just an average company. And their job description was a product that a plant manager in London, United in UK, like, we don't have any people in London, in the United Kingdom, and we don't do plant managers, you're like, someone just posted on our site. That was clearly wrong. And now like, how do we improve that? Because if someone else goes and sees that, you know, so we're not trying to bait it. Like, we disagree with your opinion, you're like, clearly it's just not right. So we want to find a way to clean up but I would want to know, where where else are we being like incorrectly? Reviewed right? Whether it's yeah, now


Joe Kurian 53:51

I think in the future, something like my FICO, right, you get sign up a fee and do it. I think that is, that's a good idea for the future. It really.


Sony Gabriel 54:02

Now one thing like you brought an interesting point. So when we look at the score that we built, right, like your was your mentioned earlier, 90 person is very critical to news. There's a reason behind the news. This, we feel like news system more balanced, medium out there in terms of providing information, because reviews, like you said, you wouldn't find too many people going to any review site to give you a positive rating. It's heavily skewed and in terms of you go there to usually to complain or to race and unsavory experience. Right, right. Right. When it comes to social media, it's always you have to take with a pinch of salt. It's always there's a ton. There's a lot of noise in there. Right? So different platforms have we have seen different trends, we platforms tend to be very negative. And we see that in the score that we have specifically for that platform. So every platform has got its own score. And then we have the aggregate VSI. The BSI is a little more balanced because of the news element in there. So that's also something that I just wanted to bring up.


Kevin Carney 55:11

Yeah, that makes sense. I, I like to think I do my my small role in pushing back on that negative sentiment, I only only post when I have a very good experience. Because yeah, everyone just likes to complain on social media. There's a lot of good work being done out there. So you mentioned community banks, and as a target client, right? Because they are, it's probably easier to go talk to the right person, less bureaucracy. And they probably need to know, like, he's mentioned relationship banking, their companies are harder to find. And so that makes more sense for them. Are you working with some of the the core banking platforms in that market space, and trying to bolt on something? So whether it's FIS or Jack Henry, or q2 things along those lines?


Sony Gabriel 56:11

Yeah, we do have that in the roadmap. We have, we have something going on with FIS at the moment. We had some initial discussions with Jack Henry, but, you know, we posted for the time being, q2 is also something that we are interested in. So we are trying to see see, as a FinTech startups, sometimes it's easier to go on the back of established from, like an FIS or a q2, because it becomes easier. You know, there are multiple ways to balk your solution into what they have, and offer them and, and for, it's assuming that the other sites also has a benefit in doing that, that you have a commercial lending suite of applications, we can bring something of value to add to the commercial lending suite of applications. And similar with different platforms that we talk to, but for us, it's one, once we establish the credibility with that provider, it's an easy, easier way for us to reach a larger client base than trying to go all out alone, right? Because a lot of the things like the vendor approval and all those things are taken care of. So it's something definitely we are interested in we are it's in our roadmap, and we are working towards it. Okay.


Kevin Carney 57:31

So you guys, you guys have to go through the vendor management process and the third party risk, like, hey, we can help you with this, we can we can do our own score, we'll give you our own score. And you can provide scores while you're there. So maybe you get a two for one winter selling to the credit portfolio management team, then you can also sell to the vendor management once you get brought on board. Alright, two for one.


Joe Kurian 57:54

Yeah, good that you mentioned it Kevin, in one of the banks, when they gave us the vendors to monitor our name was also there on the list.


Kevin Carney 58:07

This is where to start putting lots of good Yelp reviews for years. Alright, so so you're just kind of wrapping up what's what's in store for the next 12 months of fundraising, the product announcements, client growth metrics, like what's what's what's next 24 months look like? Sorry? Well, that's


Sony Gabriel 58:29

more of like, client growth, expanding to more community banks and credit unions. While we have clients in India and Germany, those came by chance. Those were leads that came to us, or we had some connections, which got us those leads. But otherwise, our focus is on in the community banking and credit union side in the US, and all the walking wants some of these partnerships with the likes of FIS, because it gives us again, trying to reach to this similar segment, right. So product market fit, establishing that a bit more within some of the segments that we talked about is the roadmap for the next 12 months, onboard maybe another 15 to 20 clients during this period, and then get the million in ARR. That's our immediate short term. Okay. Yeah,


Kevin Carney 59:26

it's a great goal.


Joe Kurian 59:28

Yeah, I just want to add one more on the product development side. What, what we put on hold because of the, the funding issues and other things, but something that in our roadmap is, you know, ever since we got into you know, FinTech sandbox, you know, we had a partnership with experience and the experience had, you know, they given us like 100,000 businesses, you know, historic information, right, this data extracts from the history And we want to use that and, you know, redevelop the product, you know, into, say, for example, if we can, you know, develop the product in a much better way blending bureau data, and also make it more, more predictable. ESCO. Right. So that is in our product roadmap, and, you know, it's on hold right now. But, you know, soon, sooner, we are going to work on it. And we already have the data from experience, but we couldn't use it yet. So that is next in the roadmap for us.


Kevin Carney 1:00:35

Okay. Yeah, get a few more clients as the product gets more clients and the product. Yeah, keep going back and forth. Great. Do you have fundraising coming up?


Sony Gabriel 1:00:47

The immediate focus is to get to that 1 million arr. Okay. I think other things will naturally follow. It's probably not the best climate to raise funds.


Kevin Carney 1:00:59

Well, yeah, that's true. That's that's a great point. Yes. Hopefully, we're coming out of that soon, as is values have have stabilized and gone down. They've stabilized as inflation and interest rates have stabilized. Hopefully all this comes back.


Sony Gabriel 1:01:18

Right? Yeah. Okay.


Kevin Carney 1:01:20

Great. So if someone wants to get in touch with you, you know, beyond the the website, are you guys going to be in any sort of incubators or events that you're speaking at? Or like, what? What is where can somebody see you if they want to see you or reach out to you


Sony Gabriel 1:01:44

about accelerators and incubators, we don't have any immediate plans to join, join any of them at the moment. But we definitely might look at the ICBA think tech accelerator program, because that's a great way to reach out to community banks. It's something that we are looking to apply. Otherwise, website, LinkedIn, we routinely have blogs coming out. So at least one or two blogs a month, once in two months, once in one we published case studies. So that's a good avenue. We have a YouTube channel where we publish content, events, I think, I think the next big event is probably sometime in September or October, that kind of timeframe. So maybe at that time, we probably would be looking at attending some of the events.


Kevin Carney 1:02:38

Yeah, I would think that varies organization or the chapter he's involved in, in Charlotte, right, the global association of risk professionals. That's a good a good group to get involved with, I would think, yes.


Joe Kurian 1:02:54

We are looking into the community bank associations in various states, starting with California, you know, Chicago, that region, so that it's also kind of close to us founders so we can travel easily. And so and Vegas is nearby. So we're looking into various events that are coming up. So we can present ourselves there.


Sony Gabriel 1:03:22

Good way, good. excuse for us to go back to Charlotte and meet with revtech laughs Yes.


Kevin Carney 1:03:28

Yeah. Well, I'd rather go to Vegas. How about that? Vegas? Well, I have a good time. What Sony Joe, thank you so much for joining us. Really appreciate it. I think you guys have a great story. Great product. I love the fact like I said it's additive to the credit risk management process. I love the fact that you guys been doing this for longer than anyone else that I've been speaking with. Because you guys were kind of the leaders before became cool. So yeah, best of luck to you guys. I'm really, really impressed.


Sony Gabriel 1:04:02

Thank you so much. Thanks for having us. Been a pleasure talking to you. And good luck to Kingsmen software and everything that you do.


Joe Kurian 1:04:11

Right yeah, thanks Thanksgiving given for having us. You know, it costs money. And, you know, anything, no partnership with Kingsmen software. We are open for all that data. Thank you. Thank you so much for this time.


Kevin Carney 1:04:24

Awesome. Thanks.


Sony Gabriel 1:04:26

Thank you

Learn about Kingsmen
Contact Us