The Connector.
The Connector.
The Connector Podcast - Nordic Fintech Week - Transforming Fraud Detection and Credit Risk: Provenir’s AI-Driven Platform with Sophia
Unlock the future of fintech with insights from Sophia of Provenir, live from the Nordic Fintech Week in Copenhagen! Discover how a 20-year-old platform is revolutionizing fraud detection, credit risk assessment, and customer management—all in one seamless interface. Sophia not only shares the impressive history and capabilities of Provenir but also dives into the transformative role of AI in enhancing decision-making processes and providing critical insights into customer demographics.
Curious about how AI can optimize your business strategies or how to effectively manage KYC, AML, and fraud risks? This episode covers it all, including the fascinating challenges posed by deepfakes in verifying customer identities. Tune in for a comprehensive look at how Provenir’s platform serves as a central hub, integrating traditional banking systems and compliance protocols to offer a holistic view of customer creditworthiness. If you’re passionate about the intersection of AI, fintech, and customer management, this is an episode you won’t want to miss!
Thank you for tuning into our podcast about global trends in the FinTech industry.
Check out our podcast channel.
Learn more about The Connector.
Follow us on LinkedIn.
Cheers
Koen Vanderhoydonk
koen.vanderhoydonk@jointheconnector.com
#FinTech #RegTech #Scaleup #WealthTech
Welcome to the Connector Podcast, an ongoing conversation connecting fintechs, banks and regulators worldwide. Join CEO and founder Cohen van der Hooydonk as you learn more about the latest available trends and solutions in the markets.
Speaker 2:Welcome again to another podcast at the Nordic Fintech Week, and with me Sophia. Sophia, I believe you come from London. I do indeed, yes, so what brings you here to Copenhagen?
Speaker 3:Well, today I am here to meet new people and make new connections and maybe make new friends, and to learn a little bit about what's happening in the Nordic fintech environment and also learn a little bit more about what our customers might want from our product.
Speaker 2:Excellent. You're obviously here to represent a company.
Speaker 3:Yep.
Speaker 2:Which is the company that you represent and what do you do?
Speaker 3:Yep. So the company is called Provenir and their business has been around about 20 years or so, so really established global business and we're in fraud and financial crime and credit risk, customer management and collections. So it's a platform that allows you to do all of those things on one platform. That's amazing. That's a lot of different things. Do all of those things on one platform.
Speaker 2:That's amazing. That's a lot of different things.
Speaker 3:If I may say yes, yeah.
Speaker 2:Now I know that you guys are also sort of involved better with AI technologies.
Speaker 3:Absolutely yes, Can you?
Speaker 2:tell a little bit more how that has been used today.
Speaker 3:Yeah, so our platform is used for orchestration of different data sources.
Speaker 3:So our platform is used for orchestration of different data sources, but also to then take those orchestrated data sources and make decisions from them so you can build a decision flow. But what we want to help our customers do of all of the different data that's coming through that system. So we use AI to optimize those decisions and to help customers get their most efficient routes, but also, on the later side, to use AI to understand a little bit more about what's happening with their customers and maybe help them to get more of a view of their customer demographics and where they should go from a strategic perspective as well.
Speaker 2:Would you say that the decision taking part, it's the glue between the different product lines that you have.
Speaker 3:Yeah, absolutely so, bringing together all of the different data sources. So, for example, when you're onboarding a new customer, you want to understand their credit worthiness you want to understand their kyc, their fraud risk and also potentially some aml elements as well.
Speaker 3:So peps and sanctions, that kind of thing, and the glue part is to be able to bring together all of those different elements and get a holistic view about that customer so you can provide the best service for them. So which are going to be the best products for them, at what credit rating, what pricing, and also, if they have any fraud risks that come up, how are you going to handle that through their customer journey? So you're going to put a flag on that and you're going to monitor it. Are they maybe at risk of being a fraud victim?
Speaker 3:So you want to be, able to monitor that over time. And then, when it comes to the later part so maybe they've gone delinquent or something's gone wrong on that credit side you're going to look at the collections and how best to reach out to them and how best to get the best, you know, best outcome in that situation would you then somehow sit between the traditional product systems or the core banking systems or the compliance systems?
Speaker 2:and I think on the other hand, together, yes, and on the other hand, probably the CRMs, where you actually have touch points with your clients.
Speaker 3:Yes, absolutely Exactly that. So what we are is the hub in between of all of those systems, bringing all that information together so you can get insights into that?
Speaker 2:And Sofia, I think when I was doing some investigation about Provenir that I read something about deepfakes. Was I correct?
Speaker 3:Absolutely yes. So when you're onboarding a new customer, typically you don't have any information about them. So what's going to happen is you need to be able to validate their identity and verify that information, and so typically what you're going to do is you're going to ask them for some details about themselves, and they might use identity documents, for example, or give you a proof of address, or you might need additional verification to be able to do that, and so, as part of that process, sometimes you know if they're a bad actor. For example, they might be using a stolen identity or synthetic identity or even ID documents which they've manipulated in order to you know kind of use other people's details that they might have got hold of.
Speaker 3:And so that's kind of where the deep fake is getting involved. So, with the paradigm shift in generative AI, which has happened over the last two years. What we're seeing in the market is a real increase, something like 31-fold increase. So a huge increase in using ID documents and various documentation which is produced by generative AI.
Speaker 2:Which makes it extremely easy to do, also right.
Speaker 3:Exactly so. It's extremely easy to create it. They don't need any coding experience. They're able to do this on a vast scale, Whereas before you would need to use highly skilled technical people who had image processing capabilities. And they'd be there for hours editing the images.
Speaker 2:Now, it's just a prompt and you've got it right. Yeah, yes, a few prompts To make it easy. Black and white, exactly, yeah.
Speaker 3:So, with that capability in anyone's hand, anyone who has an internet connection now has access to generative AI LLMs. You know chat GBT.
Speaker 2:And Sophia, you mentioned documents that are pretty much regulated documents like a passport or a driving license. Absolutely Does it also go further than that, like maybe a video of somebody.
Speaker 3:Yeah.
Speaker 2:I think a liveliness check would come in. But would it also be, like, I don't know, a voice, a video? You can really go very far in that sort of new world of creativeness, right.
Speaker 3:Yeah, so there's some more basic documents even so.
Speaker 3:In the UK, for example, just to validate an address, sometimes you give a proof of address so that will come from a utility bill, for example, or council tax bill. So that's kind of on the basic side of it. But as the kind of detection platforms have got more sophisticated, we're then asking customers to provide their passport. Plus, they need to do a live selfie, um. But fraudsters are kind of getting around that as well by using deep fake videos, um, so they can insert those into the kind of projection routine as well. So we have to think outside the box about how we're going to be able to detect these types of attacks.
Speaker 2:No, absolutely. And what I also liked about your offer is that you combine another magic sauce, which is low code, no code.
Speaker 3:Yes.
Speaker 2:So you allow the users actually to make decisions themselves. Absolutely yeah, how do you see the actual pickup by the users using that sort of platform? Are they still using you guys as professionals, or do you really see that they literally do it themselves?
Speaker 3:Yeah, that's a really good point, and I think a lot of businesses in our space talk about low code no code, and in my experience I've not seen that work in practice as well as kind of marketing as it could all the marketing kind of side of it.
Speaker 3:But since coming to Provenir and I've only been there for about seven months or so, so still learning as well I've been really impressed by the way that um customers can basically do maybe one phase of a project with our professional services team, who are a team of domain experts within credit risk and fraud as well, and then, because the architecture and the design of the solution has been put together that helps the customer to be able to do all of their second lot of configuration and any changes, they can just completely do that themselves.
Speaker 3:It's fully drag and drop, very intuitive. I had never seen an interface like that and I was thinking, wow, this is a great product, like just really easy to use. You can make all your changes, you can test it in the sandbox and then you can promote it to live, and it just doesn't have any of the kind of oh, you know, I'm gonna have to spend hours trying to make these changes and I'm gonna have to call up tech support and you know it doesn't have any of that. It's been really well thought through to spend hours trying to make these changes and I'm going to have to call up tech support.
Speaker 3:You know it doesn't have any of that. It's been really well thought through and I think part of that has come from the fact that we've been around for a long time, so it's a mature platform. But also the way that we like to do product management at Proveneer is very customer led. So we will hold forums, we'll get feedback from customers and the kind of feedback loop is very quick for that.
Speaker 2:That's the reason why you're here.
Speaker 3:Exactly, exactly why I want to talk to customers and see what they think of the solution.
Speaker 2:Brilliant yeah, there's three other words that also strike out for me and it's real-time. Yep 120 and data partners. Maybe that could be two words that also strike out for me and it's real-time 120, and data partners. Maybe that could be two words, so maybe four.
Speaker 3:Yeah.
Speaker 2:If you could go a little bit more, maybe in the 220 data partners, because that's a lot.
Speaker 3:Yes.
Speaker 2:What sort of data is in the system?
Speaker 3:Yes, so just previously I mentioned that we're looking at the applications, the onboarding, loan originations, that kind of thing. So well, what kind of data will you need? You'll need to have information about the credit worthiness of that particular customer and, depending on which country you're in, that can be acquired in different ways. So there might be credit bureaus, but in some countries there's also government bodies that provide that information. And then there's also kind of alternative ways to get that information as well, through open banking for example Based upon history.
Speaker 3:Newer types of approaches to trying to understand customers' credit worthiness. And then you've also got the KYC fraud AML side.
Speaker 3:So you want to be able to understand the customer. So does this identity exist? We can validate the name, address, the date of birth, email address, that kind of thing, but then we want to verify the customer, so then you've got to look at other details that might help you to triangulate that information, to get a sense about whether or not that is actually who they say they are. Is that the person you're talking to? So you might look at things like the non-repudiation yeah, exactly throughout the difficult world.
Speaker 3:Yeah, so device, behavioral biometrics, so information about, about how they're operating on those platforms, even social verification. So how many times have we seen that email address? Is it active? The email address. How old is that email address so we can plug into all of these different types of data partners?
Speaker 3:And because we're global, there might be different data partners for different regions. We're global there might be different data partners for different regions, so we're also going to be able to advise customers on which regions and which data is going to be best for them okay.
Speaker 2:Well, to be honest, we're almost at the end of our podcast.
Speaker 3:Time flies that's a shame. I was really enjoying this. Yeah me too.
Speaker 2:Me too, I was surprised whoop. So with all that kind of interesting information, sophia, how do they best come to you?
Speaker 3:How do they best? Oh right, so they can reach out to me on LinkedIn Excellent. Or you can go to the Proveneer website. You can ask for a demo. Or, if you're here today at the FinTech event, you can come and speak Just pass by yeah, come and speak to us at the booth.
Speaker 2:Excellent, thank you very much for joining me.
Speaker 1:Thank you also to the audience.
Speaker 2:You're welcome thank you also to the audience and stay tuned. More news from the fintech world.
Speaker 1:Thank you very much thanks for listening to another episode of the connector podcast. To connect and keep up to date with all the latest, head over to wwwjointhaconnectorcom or hit subscribe via your podcast streaming platform.