Elizabeth Whitelock

Alan Kohler: Well, Elizabeth, perhaps we should start with a description of your software. Is it a form of artificial intelligence?

Elizabeth Whitelock: It has been described as that. It certainly takes human reasoning and helps human beings get to a place where the decisions they can make are far more accurate and less biased. So, artificial intelligence is certainly a term that's been bandied around. We are described by Gartner as prescriptive analytics. And what that means is that our software allows you to take a problem, any kind of problem, typically something complicated and with lots of moving parts, and add information from any type of source including, for example, Alan, your opinion could be used as a form of expert opinion. We could take that, and we could run an assessment. And the system is designed to help us deal with uncertainty, unknowns, unreliability, it allows us to deal with missing data, and poor quality data. And...

AK: So, the problem. So, it seems to me, so what you're saying is that it, it sort of combines in in analytics, combines both objective data and subjective opinions.

EW: Spot on, yes.

AK: ... which seems to be, may be a problem because opinions aren't always correct.

EW: But if you don't have anything else, what do you use? So, as human beings, we make decisions and a lot of them are made by our gut feel, you would probably agree. And even though we say, "Oh, this is the right reason for making this decision," the reality is there is a whole different reason behind it. So, there's the right reason and the real reason. Human beings are emotional, and despite the vast quantity of data that we have surrounding us today, the reality is that organisations that we're dealing with still can't make sense of that data. So, they're swimming in data, in fact probably drowning in data, how did they get to understand the answers that they're looking for. So, what we tend to come at, we come at this in a different way. We come at it from a problem centric perspective. So, we start by understanding the problem, not understanding the data. So, instead of drowning in the data and chasing rabbits down particular holes in search of answers, if we come at it and say, "What are all the moving parts that are going to help us identify the answer," then you might go look for the information, not in all the data that you have, but you might turn around and say, "Actually, we do need to talk to Alan. He'll have an opinion on this because he's an expert in this financial market space. And maybe we should go talk to the chief economist at Commbank because he'll have an opinion as well. Because we don't have data." So, when you've got nothing, maybe the opinion is as good as you're going to get.

AK: So, before we get onto the business model, I just want to talk a little about the background. The software came out of a cooperative research center called the Distributed Systems Technology Center...

EW: That's right.

AK: ... which actually, which actually folded in 2005. Now, the rule was that IP that came out of CRC's was auctioned to the shareholders of the CRC, so who got this one?

EW: So, the, our chairman and his partner. So Rick Anstey and Laurie Hammond had a fund called iQFund and they bid for, for some of the IP from the CRC and they walked away with, as they describe it, a box that had a couple of disks and some paper in it. So, they had, I think, around six projects in there that the CRC had been working on. Some of them went on to do bits and pieces, some of them didn't do anything at all. And this product, as it was called then was Sheba, as in the Queen of Sheba. So, they had that product and they started to dig around and try to understand it. The product had been built for defence and so they went back into defence to explore further what could be done with it. And for a period of time, it really didn't do anything. They had it in incubation until the time was right. And so...

AK: And that would have been, that time was right, would have been 2010.

EW: 2010, that's right. It came off ice and one of the original developers was brought on-board. Rick and I had worked together, so, in the past, so her asked me to have a look at the product to see what we could do with it. And here we are today, so it's been an interesting journey to this point.

AK: And, the contract with the Department of Defense began in 2012. So, how much are they paying and what are they actually paying for?

EW: So, the Defence Department signed a contract at the end of 2012 and then we sorted the role out in 2013. The agency that it's deployed into in that particular contract is the intelligence organisation. And the intelligence analysts use the software to help them assess threat events, likelihood of certain elements, people perpetrating particular scenarios, shall we say. [laughs]

AK: [laughs]

EW: Sorry, that was very roundabout, wasn't it?

AK: No, it was just, I imagine you can't be too direct about what you're doing for the Department of Defense, I suppose.

EW: No. No. [laughs] Thank you. No, we can't.

AK: Okay, well, let's talk about Tyndall Capital, which is another of your clients...

EW: Sure.

AK: .... and maybe the best way to deal with, to talk about your business model, is what you do for particular companies and how they go about paying you.

EW: Sure.

AK: So, from what I understand, Tyndall is using the software to work out the likelihood of repayments, so they can work out the pricing of loans. Is that correct? So, your...

EW: Yeah, in effect.

AK: ... your algorithm kind of takes the assessment of that data a bit further.

EW: Yes, so in the Marketlend scenario as a peer to peer lending platform or digital lending platform, their focus is small to medium enterprises, and Leo had a lot of experience in the lending world, so what he was aware of was the downside of some of the credit risk assessments that would be performed, that they weren't necessarily always truly reflective of a person or an organisation's ability to repay their loan. So, he looked to our software and we, with him, we created a credit risk engine for want of a better description. So, our engine with its algorithms in it perform an assessment on the likelihood and certainty of a particular company repaying their loan. And in that process, the Marketlend system gathers information from the applicants, that information, there's several things that are checked along the way to outside agencies, including things like the ABS, and then information is sent into the engine and an assessment is performed. And it comes back and says, "This is how likely a company is to repay the principal and interest, here is the confidence rating." If it hits a certain benchmark, then it goes to the market for lending. And...

AK: And, so I'm just, sorry to cut you off, but you've got another one. You've got another contract which involves marketing analysis, similarly, in a way. And also another one that's quite interesting which predicts the failure rate of spare parts.

EW: Yes. So Oniqua software, they are a company that originally came out of Brisbane. They have sixteen billion dollars-worth of assets under management in their systems. So, their clients of the big end of town typically, the BHP's and Rio Tinto's. And they're trying to optimise the inventory of those organisations so there's no opportunity for production line to come down, so they have built a model using our software that allows them to asses every single inventory part that sits inside BHP's, for example, database of inventory. And it produces a criticality rating, so how critical is this part in the production line, and is this part something that we can replace easily. So, is it just a general consumable that I can get next door, or is this a custom made part that's going to take four weeks and has to be flown in. So, trying to avoid, for example, if you remember the Olympic Dam scenario, last year was it? Where Olympic Dam was done on weeks on end, and you can imagine how much that would have cost in terms of lost production. So, it's making that production line stay up and that you have the right amount of inventory sitting there in your stockpile ready and available to apply should anything happen.

AK: So, how are these customers paying you? And how much? I mean just so I could get a sense, I understand there are two ways of you being paid.

EW: So, there, so there's, we don't discount the notion of being a direct client. So, for example, in our public safety arena, so that includes defence, law enforcement, counter-terrorism, they are typically clients who pay us directly. So, there's a licence fee for the use of our software and then we would train them on how to build models inside their air gapped or secret networks. So, there's that relationship with is just a typical software company relationship with their client, provide software and services. And then with Market Lend, and Oniqua, and some of the others that we're currently working with, it's a partnership. So people like Oniqua have a group of clients that they already have sold to, and there will be new prospects. And they're looking to add value to those clients. We don't know the mining space. We know our software really well, but they understand the problems of their clients. And so we work with them to build the right algorithm to solve the solution that they're looking for and we share in revenue. So, we get a clip of the ticket if you like. So every time someone performs an assessment, say in Marketlend, there's a clip of the ticket the comes into the Veriluma bank account. With Oniqua, it's a sharing of revenue. And the same with RPMG and our joint venture with Legal Logics in the legal sector. It's exactly the same, a fifty-fifty revenue split.

AK: So, what are you saying is the growth potential for revenue here? What, what, I mean, how big is the pond that you're playing in?

EW: Well, this is an application that is global. We don't, we are, we enable so many different problems to be tackled. So, this is a global application and we already know that it can be applied to many different industries. The Gartner group say that prescriptive analytics specifically, which is the ability to predict what might happen with a view of all the risks, opportunities, and a path forward - that market is the next evolution in the analytics space. In the next three years they believe there will be a twenty-two percent compound annual growth and that by 2019, the revenues into prescriptive analytics will be around one point one billion. So, it a big pond, it is on the up, and right now we, in terms of the players that are there, that are players who are doing very specific niche focused application, whereas we're coming at this with, in effect, like a big box of lego. Our engine allows you to create solutions that fit your particular problem space. And it allows you to then deploy that within the existing infrastructure so you never have to see our application in effect, it can run headless in the background of whatever you have. So, my goal, and my personal view, and I know that I'm stealing someone else's tag line here, so my apologies, but Veriluma becomes the intel inside, the intelligence inside other people's processes, systems, technologies, and applications. And if you go to Marketlend, or you go to Oniqua, or any of the other partners, you won't see Veriluma there, you won't see our interface, but what you will get is an assessment, and you will get an assessment of your risk, and you will understand exactly what has contributed to that assessment right here, right now.

AK: Just finally, one of the interesting applications is in family law, you've got a fifty-fifty joint venture with Templetons.

EW: Yes.

AK: Could you explain how your algorithm is being used in family law?

EW: Sure can. So, we are completing development of this right now and if you look at the family law courts and the system currently in play, you'll know how, to quote someone else, "how constipated the system is". There are close to fifty thousand divorces a year, approximately eighty percent of people are self represented, so they're not using lawyers. And Ma and Pa don't actually understand the law and what they're entitled to. Our application is designed to provide an interface whereby they can enter details of their situation, their children, and so on, and then our engine is performing an assessment to tell them, "Here's an indication of what you're likely to get as a property settlement." So, it gives people a view of their property settlement and all of the elements that are contributing to that particular split of the assets, which means they can then choose to print a brief and use that brief to help conduct the self representation part, or they can choose to talk to a lawyer that we will have sitting behind the application. But the goal here is to actually give the consumer something that allows them to take control of their situation, to empower them, to walk through the legal system that is very complicated and actually understand where they are positioned and how to tackle that.

AK: So, I presume this is a sort of a retail product that you'll market with Templetons. Is that right?

EW: Yes. So, in the Australian market place, we are very much focused on the consumer. It does not prohibit, you know, lawyers potentially, say, using it for their clients and getting an assessment by their client filling in the details. But our initial go to market here in Australia is certainly consumer based. We have had interest from other countries and geographies where it might take a slightly different spin.

AK: So, have you worked out yet what your pricing will be?

EW: So, rough pricing. We've spoken to a number of focus groups around this and given that a lawyer spend is probably going to cost six to eight hundred bucks an hour, initially for using the application in getting an assessment, we reckon that in bite sized chunks in that process, we could probably charge in total about two hundred dollars to use the application. And you'll come out at the end of that with an assessment of your situation and a brief that allows you to then move forward and into the court system.

AK: I'm presuming that would be almost all margin.

EW: Well, it's costing us to build.

AK: Yes.

EW: So, actually, there's development effort around the model. There's development effort around the front end. We have to host the application. We have to have security. We have to have payment gateways. We have to have access to barristers and the Law Society to make sure that everything is above board and isn't going to upset anybody. So, there's a lot of effort and a lot of money right now being ploughed into this. So, it's not going to be all margin for a while.

AK: We'll have to leave it there. It's been great talking to you, Elizabeth. Thanks very much.

EW: Pleasure. Thank you so much.