Podcast: Gen AI can provide FIs up to 90% automation

Vaidik Trivedi 0:02
Hello, and welcome to The Buzz, a bank automation news podcast. My name is Vaidik Trivedi and I’m the associate editor of bank automation News. Today is January 9 2024. Joining me is Adam Famularo. He is the chief executive officer of workfusion, an automated solutions company. Adam has served as the chairman on the board of directors of lucid work prior to joining workfusion, and brings decades of experience in automation and digital transformation to work fusion. He’s here to discuss how financial institutions can leverage AI driven digital workers for efficiencies, and his company’s plans for 2024.

Adam Famularo 0:44
You got it, buddy, Ken, thanks for having me. So I’m Adam Famularo on the CEO of workfusion. You know, I’ve been in the software industry for the last 25 years as CEO, General Manager, all the way down to individual contributor roles. So I’ve been doing this for quite a while. workfusion is by far the most exciting company that I have worked for to date. Port fusion workfusion is a mixture of three really trendsetting technologies with RPA IDP, intelligent document processing, and AI and machine learning. And our technology has been around for about 10 years now. It was born out of MIT labs. And it’s been proven out really over the last, I would say seven to eight years within large enterprises. We started as kind of a holistic platform, we call it work.ai platform that can in essence, enable any company to take any complex data source, and add an AI machine learning algorithm to it in essence, a brain that can create and cause action on top of what it just read. And then from there, it can then build it, we have a learning mechanism that allows it to learn and develop and grow. About two years ago now it’s actually two years, February 2 2022, we launched what we call digital workers. These are our fully packaged workforce, right. So it’s an AI powered workforce, that allows companies specifically financial crimes, focusing on financial crime, so doing job roles, like sanctions and sanctions, screenings, and job roles in the whole KYC process of selling targeted to financial services companies. So we built out these digital workers. And now we’ve we have, in essence, built a large business focusing in on the financial services companies, we still do provide our work.ai platform for other companies and other industries that want to build or develop their own custom digital workers. So that our business

Vaidik Trivedi 3:09
Alright, so digital work is that sounds very interesting. Let’s talk a bit more about that. Tell me how is it developed? And what are some of the best use cases of these digital workers in financial services industry? Yeah,

Adam Famularo 3:24
so what we’ve done is we’ve taken really a honed in focus on IDP. So So the notion of being able to read or decipher documents, and in working with the top banks in the world that were already our customers using our work AI platform, we took the best machine learning models that they were already using today, and package them up as actual digital workers. The reason why we did this work was the banks that were successful using our software, they felt like naming them, giving them a persona made it easier for them to be digested and used within their corporation. So in studying companies and seeing what they were doing in success, we said, well, let’s take it to the next level. Let’s give them names was given faces. Let’s give them personas. Give them a specific detailed job description and job role. So that when you view our digital workers, you’re viewing them as hired workers that you bring on board as kind of copilots to go along with the people that are doing sanctions and sanctions screening or job roles in the KYC process. What we want to do is automate the remedial work, right? The work that most humans don’t want to do so very heavy data intensive reading documents, reading databases, reading websites, as part of the decision making process.

Vaidik Trivedi 4:58
Okay, Can you tell me what datasets do you use to train these AI digital workers? Is everything in public domain used to train them? Yeah,

Adam Famularo 5:08
we do. We use everything from software applications that they might be using today in the sanctions and sanctions screening process. We’ve recently launched partnerships with companies like Thomson Reuters, where we’re taking in reading and deciphering their data as part of our process. And we’re working on multiple other partnerships right now that will help us support those data flows. And then more importantly, a lot of our customers are allowing us to take the datasets that they have trained their models on, and use those datasets to share with other banks and financial services companies. So that when you hire one of our digital workers, it’s like hiring somebody that’s been trained for the last three to four years in doing that specific job role.

Vaidik Trivedi 5:58
Okay. And you said that you gave your digital workers a name of face so that you can just onboard them into your company. Can you tell our listeners about who Tara is and who Kendrick is? Yeah,

Adam Famularo 6:12
so you’re thrown out some of our names. So So Tara is a specific job function doing transaction screening, right. So what she’s doing is she’s viewing all the transactions that are coming in. And then using outside sources inside sources, she’s deciphering whether or not that that is a good invalid transaction trying to stop the bad guys. And if she finds one that looks like it is, she will then tie in a human, we have this thing called a human in the loop that we developed very early on. But it’s in essence, an algorithm that pulls in a co worker, and that co worker will then view it and say yes or no, and approve or disapprove. And then terrible kind of go about her job. So that is our transaction screener. Then the other men, the other person that you mentioned, was Kendrick. So Kendrick is part of the KYC process. So he is doing Id check. So he’s checking that you are who you say you are. He’s looking at passports, he’s looking at driver’s license or any other form of ID that you might have overseas, and then deciding whether or not that you are who you say you are as part of the KYC process. And those two are part of now seven digital workers that we have in the sanction screening and KYC process.

Vaidik Trivedi 7:33
Okay. So for digital workers, can you tell me why would a financial services company want to have them? Does it bring monetary savings? As they speed up that operations process?

Adam Famularo 7:47
Yes, yes. And yes, the more. So the reason why you you bring on one of our digital workers or multiple digital workers is the first part, it’s very hard to find the people to do these job roles, these are really level one job roles. They’re very remedial work. It’s very labor intensive, and it’s very error prone. So with all of that, it’s easier to hire one of our digital workers to come to work for you, and then work alongside the rest of your team as that AI kind of co pilot example. So that’s the first part then the second part is that this is very, it’s a regulated space, right? So we’re talking about areas where if the job’s not done the right way, or it’s not documented the right way, you can be fined. And what we do is we give it a an easier shot for regulators to be able to come in and see all the decisions that were made by Tarot or Evelyn, and understand why they made those decisions behind the work that they’re doing. So it makes it easier for the regulators to do their job. And overall, we’re looking at a reduction in costs, right? So it’ll cost you less money to do the work without digital workers, you’ll get much faster time to value. And it’ll be much easier the day that a regulator comes in to want to see what the work that you’re doing and prove out that work. Were we actually documented all and it’s ready to go. All

Vaidik Trivedi 9:18
right. Can you give me a few examples of the financial institutions that are using your solutions as of now? Yeah,

Adam Famularo 9:27
so we’re in pretty much most of the top 50 banks around the world. Very heavy focus in North America and Europe. From our business standpoint, of course, most of the banks don’t like us to name names. So you know, I don’t usually tell their names. But we are across the board on all the large banks. And usually the way that they start with us is they usually start around sanctions and sanctions screening. And that’s Tara Neverland. So Most banks today, if they’re either on boarded with us or looking at us, they’re looking at us for sanction screening. And then the ones that are with us for the longest period of time, they usually then move over to the whole KYC process, where, you know, we’ve just built and developed what we call a P KYC. Digital worker for perpetual KYC. So getting to a state in point where we can monitor and manage KYC, in perpetuity. So that’s, that’s usually how we set up and work with banks today.

Vaidik Trivedi 10:33
Okay. As a for AI, it’s a very rapidly evolving field of technology, from machine learning to now we have generative AI, how are you looking towards generative AI? Do you have any use cases for this?

Adam Famularo 10:48
We do we do. We have plenty. So. So look, we’ve since we’ve been in AI for shoot over over three years now, we have a lot of built knowledge and understanding of models and developing models. The beauty to our AI for the banks is that we are a white box, right, so you know how our AI models are built, we update the AI models with our network learning on an ongoing basis. And they know and understand the power of our AI. And it’s documented and open for everybody to see, including the regulator’s with Jenny AI, you don’t have a lot of those same ingredients more of a black box behind the scenes. But we view it as part of our human in the loop process, we almost have like a Gen AI in the loop process to go along with human in the loop. So you know, we’re using it with Evelyn today when we’re doing adverse media, and looking at different adverse media as out there in the marketplace. And it’s helping drive up our automation rates where you know, we might be when we start out in implementation, we might be about 70 75% automation, with a Gen AI tool, we can get up to 90% automation. So we can, we can really kind of close the amount of automation that we’re bringing to market using Gen AI. But we have to be careful with the hallucinations that come from it and to ensure that it’s it’s part of our human loop process. So it helps build the confidence of the outputs that come from from AI software. Okay,

Vaidik Trivedi 12:25
how revolutionary Do you think Gen AI will be in anti money laundering space and financial services in general?

Adam Famularo 12:34
Look, I think in general, it’s going to impact all parts of business, right, just like I gave you the example of how we’re enhancing our software with it, I am enhancing my customer service organization with it, I’m enhancing marketing with it, I’m enhancing finance with it, I can literally enhance all the functions of my business with it. So viewing it very much through that same lens with financial services. The beauty of our software is that we help enable companies to actually prove value. Now, a lot of the financial services companies are stopping and pausing and waiting until this whole thing gets figured out because of the black box nature of Gen AI. The beauty of using our software along with Nai is that we can get you real time results today by using that software, where you can see real cost improvements and productivity improvements and doing it in a safer environment.

Vaidik Trivedi 13:31
Okay, as afford 2024. What has gotten you excited? What’s in the pipeline for cold fusion? What are the future plans?

Adam Famularo 13:39
So I started talking about what we’re doing with PK YC, that was actually led by one of our largest customers. And they said, Listen, we want to do we want to go to the next level and really make regulators happy with what we’ve been doing by introducing this notion of perpetual KYC. And we’ve just went live with them. We’ve just packaged that up as Kayleigh, we just named that digital worker and and going to market with that this year. So that one really has us going. Then the other part since we do see so many transaction data flows, were starting to cross over into fraud as well. And we’ve just launched a digital worker called Isaac that’s going to help do fraud prevention. And we’re literally just launching that one now. And we have three customers that are kind of our launching customers with it. And I believe, between those two new digital workers that we’re bringing to market that’s really going to excite our customers by getting into new spaces using AI. The other piece that I just started to mention was eight strategic partnerships. So partnership, we can’t do it alone, right? We’re still a medium sized company in the end of the day. We’re not you know these multibillion dollar company yet, but for us to get there we need to form unreel strategic partnerships. And that’s where, you know, Thomson Reuters comes in, we just did a big partnership with emphasis, which is helping us really bring to market, our financial services, digital workers to their financial services customers. So between those two partnerships, and we have about three or four more that are coming in early this year, I’m very excited to see what they’re going to do to help us really, really grow in the marketplace at a faster clip.

Vaidik Trivedi 15:29
Okay. Talking about your funding plans, and how you raise funds, in 2021. Do you have any other plans are facing any further? So

Adam Famularo 15:41
that’s a great question. And I still don’t know the pure answer to it just yet. You know, we were going to do around in 2024, late 2024, before the whole, you know, the this little bit of a mess in the financial services industry over the last 12 months. So that kind of delayed things for us, right, we just like other software companies became a little bit more cost conscious. So we watched where we’re spending our money. And we since we did do a big round in March of 21, we’ve been able to kind of push that out now into 2025. Now if there’s if there’s signs of, you know, the market rebounding, and we do see this continued growth we might do around earlier than later. But I would probably tell you, sometime q1 of 2025, more potentially in q4 of

Vaidik Trivedi 16:35
2024. Okay, as for the growth of the company, can you tell me how the company is growing, how much the revenue has grown year away?

Adam Famularo 16:45
So we don’t publish that since we’re a private company. I can tell you we are in double digit revenue growth. But I can’t share a number since we are a private company.

Vaidik Trivedi 16:56
You have been listening to the bus a backlog of machine news podcast, please follow us on LinkedIn. And as a reminder, you can leave this podcast on a platform of choice. Thank you for your time. And be sure to visit us at Bank automation news.com For more automation news,

Transcribed by https://otter.ai