In this interview, we talk to Zachary Bennett (, who is working in Business Development at ComplyAdvantage ( ComplyAdvantage is a regtech startup, which wants to revolutionize anti-money laundry. They apply AI to help banks and financial services companies to get the many regulations, sanctions and embargo lists with their proprietary database. They claim a reduction of “false positives” in sanctions and embargo monitoring of up to 84%.





Male Speaker: Welcome to start your podcast and YouTube blog covering the German startup scene with news interviews and live events

Joe : Hello and welcome everybody this is Joe from your start a podcast and YouTube blog from Germany. I’m here again on the second day in London and look at who I have stuck with me. Hey welcome,

Zachary: Good morning, how is everyone?

Joe: We should tell our viewers that it’s actually morning in London so how did you do with the [Unclear 00:00:43]  last night you joined it

Zachary: Yeah I joined for a little bit and then left quite early to be safe for today to be fresh

Joe: Okay I admittedly I was boring and I wasn’t there at all but you’re here because you’re working for a startup company right?

Zachary: I am indeed, yes I’m working for a [Unclear 00:01:00]  company called comply advantage

Joe: Now what do you guys actually do?

Zachary: And well we’re hoping to revolutionize AML screening and monitoring

Joe: For our viewers who have completely no clue at all it’s anti-money laundry. So that means there are lots of regulations out there where banks insurance company all financial services companies and a lot of other companies as well have to screen the customers their transactions and look for signs of money laundering.

Zachary: Exactly, yeah sure

Joe: And it’s a lot of work

Zachary: It is a lot of work and that’s what we found and that’s what we was trying to improve on. You know traditionally there’s been a lot of issues completing these tasks and that contains a lot of manual work which is unnecessary and it can be high false positive rates. There’s been poor integration

Joe: Your system you’re using a false positive, it spits you out okay look at this transaction, you look through it and there’s no hint of money laundry so it was a positive from the system but it was actually false so

Zachary: Yeah exactly so, that can be very infuriating you know you’re doing a lot of work and remediating these false positives and which is unnecessary and doesn’t need to be done if you have the correct system and the correct technology in place. So that’s exactly what we were hoping to do and what we have done we’ve been around for four years now and we’ve been making huge strides in solving these common issues that are found in the industry.

Joe: What are those issues? Because assume like 90 to 95 percent of the people out there have no clue how to fight money laundry and other things like what a pep is and stuff  like that.

Zachary: Yeah sure so our process is just one step in the process and with these regulations there are lots of things that you have to do and that’s kyc which is know your customer and ID verification. However, what we prefer to concentrate on is the the need to check your customers when you onboard them. And so what that means is checking them against sanctions lists, you may have seen in the news recently Donald Trump released his latest sanctions lists and there’s so many lists such as this that you need to be checking against to make sure you’re not working with anybody else.

Joe: When we talked before it was like 49 sanctions list so it’s it’s not like a list it’s websites full of content, of people of little entities and you have to bring them somehow into your system and check for all of them. If you don’t want to get in trouble.

Zachary: Yeah exactly so you’re checking against all these sanctioned people on top of that you need to be checking for politically exposed people so that can be politicians, Kings, judges,

Joe: Generals

Zachary: Yeah, Generals, army Generals of course, there’s so many people that can be influenced in this anti-money laundering fight

Joe: That means you’re basically checking their accounts if they don’t get any bribery from outside of the country if where they are. For example, a judge somewhere from a different country has an account with you and he gets like two million and then you have to consider it under those circumstances right?

Zachary: Yeah exactly and you need to be aware of this when you onboard a customer to make sure that you can keep an eye on these kinds of things that goes on. So you need to be aware, that doesn’t mean that you can’t work with a politically exposed person. It just means you need to be aware

Joe: It doesn’t mean they are all taking bribes just  means there’s a regulation you have to

Zachary: Sure and on top of that what we’ve done to enhance the offerings that were there previously is we have introduced adverse media. And we do this very well so as well as alerting you to if a sanctioned person or it can be a sanctioned company as well. And we let you know if there’s been any kind of negative news in the media be that money laundering, sexual exploitation anything that can come up in regards to the name of your customer you checked will all show up in our system on a consolidated profile. And we’ve been creating this database and it is a proprietary database there are only a few companies in the world who actually own this proprietary database for AML data.

And the difference between our data and everybody else’s in the marketplace is we no longer use human analysts as we were discussing earlier there’s lots and lots of companies and people in the world in a very volatile world that is constantly changing. And as a human analyst whether you have five hundred or a thousand it’s almost impossible to keep these profiles updated.

Joe: To get this in perspective for our viewers, basically there are whole departments of people and they sit in front of a system and they get split our transaction after people transaction people and they have to check it for money laundry, for an embargo violation, for something like that. It’s a boring job if you’re checking it all the time and nothing comes up.

Zachary:  Yeah so what we’ve done is we’ve used artificial intelligence and machine learning. So we don’t need any human analysts, all of our data is updated every single day, every risk profile you can guarantee is up-to-date and accurate which is something that wasn’t the case before in the industry.

Joe: Because you’re doing it automatically instead of manually

Zachary: Yeah exactly everything is done automatically, there is no need for a human analyst to sit there and a piece of news and then update the system. So if you put that into context with Donald Trump’s latest sanctions last week with Iran, we had all of our information updated in our system within 15 minutes.

Joe: Can you give our viewers like an impression how much this is, was it like four lines of text or was it like hundred pages of…

Zachary: Yeah exactly so it’s a big updated document basically and that has to be scoured so if you take our competitors, some of the traditional names in the industry it took them between a week and two weeks and we had it done in 15 minutes.

Joe: That’s pretty good and then after those 15 minutes was it immediately available for your clients?

Zachary: Yeah exactly so we’re cloud base. And everything as soon as it’s updated on our system is immediately available for our clients

Joe: What I also found interesting because I have here some of the papers you had in your booth and actually you’ve done a case study with one bank and the the false positives we talked about before went down like 84 percent.

Zachary: Yeah we’ve been well known to reduce the false positives quite by a high number which means less work and you know you can focus more on other sides of your business and grow from there. So the way that we would use the false positives is to make our data highly configurable. So like we were discussing if you are working in mainland Europe you don’t need to screen against APACK or Asian sanction lists or political exposed people because you know they aren’t your customers and all that does is provide a greater risk and a greater pool to to pick out names that were incorrect and false positives. So what we do is, we allow you to configure depending on a company’s risk-based approach and how wide or thin they want to do and what jurisdictions they work in.

Joe: Oh just a quick story on that because especially when you change between languages, between alphabets, between ways of writing, there’s always confusion on how to spell a name. So basically that’s a big headache for those analysts. I remember the ruler of Syria and his family name he was

Zachary: Al Bashere

Joe: Yeah and you could spell it like 60 different way The Economist wrote and you have to keep it in mind it’s really tough work and if you just miss one of the Spelling’s your bank can be in big trouble right.

Zachary: Yeah it’s really interesting that you say that because obviously these fraudsters and bad people as we like to say are they’re very cunning and there will be certain ways that they change their name or there’ll be Ali Isis

Joe: They use legal entities right

Zachary: Yeah exactly

Joe: It’s not that those people are dumb and go there with their ID card they use like fifty five six different shell companies between them and the actual company that conducts the illegal business so very interesting area

Zachary: Yeah and what we do to counteract  that is we do this in two ways, so we consolidate any known nicknames or difference in names into one single profile. Whereas previously you may have found twelve different profiles for one person depending on how many names they had so what we do is we pull all those together into one consolidated profile so you can just have one strict view. And for each name that you search will bring up the same one profile, on the other side of that you mentioned different spellings and the different nicknames and sometimes people can misspell their name on purpose knowing that they would come up on a certain list. So in order to kind of remove the issue of this and you missing these people we have a thing called fuzzy logic.

So what that means is there is a certain amount of leeway so for example, my name being Zach my birth name is Zachary. So if you had a higher fuzzy logic and I typed in Zach but he was looking for Zachary Bennet I would still come up under a nickname so that’s one way that we make sure that you’re accurate and safe.

Joe: First time you talk to me about fuzzy logic I thought he you give your computer’s a little bit of whiskey

Zachary: Yeah aha and first the hair

Joe: Oh great you’ll find [Unclear 00:11:18]  profile and the link for the company website down here in the show notes. So just have a look if you’re interested. Thank you very much was a pleasure having you here.

Zcahry: Yeah thank you very much for having me.

Male Speaker: That’s all folks find more news, streams, events and interviews at Remeber sharing is caring.


(End of Audio)