The Exchanges

Q That's awesome. Okay, so tell us why you launched this. I mean, how did you discover this problem?

A Um, well, so I actually ran a, uh, a consulting company for a decade, and in that time period, um, and before that, I was a portfolio manager at American Century Investment Management, um, where I had, um, started using recursive partitioning in addition to kind of your traditional regression analysis for, for data, you know, for data analysis, and what I found was that while regression was really good at kind of guessing what the, What the average expectation would be recursive partitioning was really good at guessing the extremes. And so, you know, when you ask, like, what are the describe your product in five words, we forecast extreme events. That's what we do. And that's what, quite frankly, that's when we lose money as portfolio managers. And that's when we get hired as investment managers or wealth managers. And so, You know, I started, um, crafting this product in the form of a consulting product about a decade ago, so I spent, um, you know, once I had lifted out of Rubini Global Economics, where I had launched a consultancy that, um, was profitable, we were able to adequately spin that out. I earned him out over three years, and in that time period, I kind of gave myself the agenda and latitude to build the kind of research agenda I wanted to build, and with my client base, Um, really focused on, on this technology and effectively created the prototype for what we've b…

AI assessment note: “recursive partitioning was really good at guessing the extremes. And so... we forecast extreme events.”

Q That's wild. Okay. I want to come back to the idea of recursion partitioning here in a second, but first to get going in 2021, you mentioned a seed round. I think you said in 2022, what was the size of that round?

A So, um, so we actually raised, uh, 425,000 dollars in 2021, uh, sorry, in 2022, uh, from 20 45 ventures, Ulu Ventures, and a handful of, uh, of individuals, and we're, um, we added to that actually early this year, um, and so we're just finalizing the second half of that seed round, which it's been the prolonged seed round. Anybody who's been raising money knows how painful it's been. To find, uh, and, and secure pre-seed funding, but we, I think one of the things that have kept us in front of, um, the, the early stage venture companies is that our product is the exact kind of product you need when markets are volatile. You couldn't get more volatile markets in the last few years. So, and strangely, you know, we were really concerned that, that the meltdown in Silicon Valley Bank was going to be, um, was going to be sort of, you know, a real death blow. Uh, to our funding process, but actually our funding set up after that, probably because people actually came to us saying, hey, I know you're not fully built yet, but can we subscribe to whatever you have? And so we're actually pre-selling, um, just access to the data engine to access to the analysis engine. And so, you know, I think that's probably what has kept us in the game while other, uh, other early stage companies are, are having troubles.

AI assessment note: “we actually raised, uh, 425,000 dollars in 2021, uh, sorry, in 2022”

Q 10 days ago, something around there. Gina, I mean, that is the definition, I would say, of, you know, recursion partitioning and extreme events. So people are going to be wondering, well, how, how, if, if no one else in the world can predict these things, how is Gina saying she could predict it? So how would your technology have identified and told portfolio managers about SVB ahead of time?

A Well, it's not going to tell portfolio managers about SVB. What it tells portfolio managers is how their portfolios respond to different, um, uh, to different Economic events, um, and what happens when they happen in succession, right? So while people understand how their portfolio might act when interest rates rise, what we do is we help them understand what is the exact set of conditions that will lead to the worst performance that they will have, and so what combination of events, and that's really where we don't have to predict the event. We don't have to predict that the pandemic will happen. We just simply have to predict that Oil prices will go up, um, that interest rates will go up, and that the Fed balance sheet will start to retrench. That tends to be a death blow to a number of portfolios. It doesn't matter what caused it, and so I think a lot of people try to focus on the crystal ballishness of it. We're not in the business of focusing on the crystal ball. We're just basically trying to say, whatever event happens, if you get a series of events, this is the combination that you want to be aware of, right? On the flip side, if you're investing for growth, You know, so if we're looking at deal by deal, uh, investments into a venture portfolio or PE portfolio, we look for the series of, of attributes or aspects that will lead to outsized performance. And so, you know, …

AI assessment note: “Well, it's not going to tell portfolio managers about SVB.”

Q can happen, and if so, how do you, yeah, so like, I mean, but how can you possibly, so let's say tomorrow I'm making this up, a Russian submarine hits the internet cable line in the Atlantic Ocean and cuts internet communications off between, you know, Europe and the U.S. I mean, how can you possibly think of all the crazy things that could happen tomorrow across geopolitics, economies, everything?

A Yeah, but let's take that crazy example that you just put out. Well, what would naturally happen? You'd have a natural A fall in e-commerce almost immediately. That fall in e-commerce is going to result in a fall in, in at least a temporary fall, um, in profitability. Depending on how long that outage happens, uh, that temporary fall in productivity could actually manifest, um, into other, um, elements like a fall in, in, uh, in labor. So you could see labor cuts, right? You could see wages. And we can tell you how your portfolio will act in those. So it isn't about guessing the event. It's about guessing how the event is going to evolve. Um, in terms of elements that can be predicted, right?

AI assessment note: “it isn't about guessing the event. It's about guessing how the event is going to evolve”