AI Fuels Finance’s (Much Needed) Tony Stark Moment

Is AI Taking Over?

AI, from the boardroom to Hollywood to your multiplex to frankly your kids’ college essays, it’s everywhere. It’s a real thing. It saved America from apparently the great recession part 2. 

But how real is it really? I’m with Joseph Drambarean and Brett Turner with Trovata. Now Brett, you’re a CEO. Not only that, but you were at Amazon the last time we saw this movie, which was basically in the form of the internet 20 years ago. Talk to me about how this is similar and how real should we be considering this AI moment?

It’s a huge moment. Everything exploding, the summer of AI, it reminds me of the early days of the internet. It really is that profound. My start of my career in the mid to late nineties, we’re going through that whole process and just seeing the explosion of companies and just even all of the older industries, banking, telecom, how is it going to react with all of this? 

And it was crazy. And what we’re seeing though, that is also, like the internet, is that it’s getting a lot of noise before it’s getting a lot of real application, at least as far as people can see in the public. 1996, Mission: Impossible comes out, the NOC List is going to leak on the internet, and that’s a big scary thing that everyone has to be worried about. Dead Reckoning Mission: Impossible Part 8 million comes out this month, AI is literally the MacGuffin, it is of the moment to that degree, but is the money really being put to work and what kind of things are happening that wouldn’t otherwise in a world before AI like last summer or three weeks ago? 

Well, it’s coming along pretty fast.

Well, first off, we got to be careful because I haven’t seen Mission: Impossible yet. No spoilers here. I’m going to be going this weekend.

Just an internet spoiler. It’s really good. Nothing happens.

My wife’s actually watching it right now. You look at all the fear that happens. I think that’s a sign that something is pretty disruptive happening. And then even yesterday hearing, reading the news that the FTC is now going to investigate OpenAI.

Oh, good. The government’s here.

I’m sure, the lead investigator, I hear, is being appointed, a guy named Bard. 

But other than that, but you look at all these things, it’s just seemed crazy. And they’re happening now every week and it’s accelerating and I think that’s just a sign, you see that underlying fear. 

And then when you look at a profession like finance, then finance folks are professional risk managers. So when you look at that and the fear permeates and hits corporations and things where security fraud, there’s already huge concerns. 

Cybersecurity has been such a huge topic, rightly so. Then when this hits, then that fear starts to set in on the corporate side as well. These are going to be, how do you navigate through that? 

These are important topics. Unless you can satisfy those things, it’s really going to be a hindrance to really leveraging AI, which there’s some really positive things of why this is the moment for finance, given this whole moment of leverage and automation that’s been waiting, I think, for a long time, for the last decade, where it’s hit IT and all these other aspects, it still hasn’t really hit banking and finance and it needs to. 

So this is a clash of riptides are starting to happen with this. It’s really exciting, but it’s also a little crazy to see the headlines.

Let me dig a little deeper into the finance point because that’s a classic sector where they were reluctant to get on the internet. They went on the internet, okay, you can look up your accounts online, and they froze there and we’ve been stuck in that same, you can look your account up online, but we don’t really want to be part of this all that much. Now we have this huge flood of capital. How’s that going to turn out? How’s this play out in terms of pushing finance over the goal line?

Yeah, I think when you look at the early days of banking or online banking, it all came out of the, once the internet hit, I think all the analysts turned to the banks and said, “Hey, you got to get on the internet. You got to be like a media company.”

Well, everyone did. Again, that’s what’s happening today. CEOs across the country are saying, “We need some AI.”

Yeah, totally.

Here comes a hundred million dollars, applied to what exactly? We’ll figure it out when we get there.

And it seems so silly, but I think through that though, there’s always these real practical applications that truly matter. 

And this is what’s fun because we’re less about trying to get headlines and raise lots of money, but we’re actually really solving problems and putting this to work in a way that’s going to hopefully drive down, and I think drive down the promise, all of this manual work that’s very clunky and you have so many things that are happening right now where you can’t get leverage in finance, you still have to do these sorts of things. It’s just required. 

Whether you’re in accounting, closing the books, there’s a process that’s clunky. Whether you’re in finance and having to do scenario planning because COVID hits and it’s… impossible. 

Just to get a team of analysts to go through that. You have to do that just through sheer volume of work with manpower. And then you look at treasury, these repetitive tasks that have been happening over and over and over and over, and automation could really drive that massively or change that massively. So it’s just waiting as a table set for these moments to happen and they’re going to happen. It’s just waiting through it.

Outstanding. Hey, it’s coming to your multiplex, it saved the world. It’s AI, folks. If you want to learn more about it, check us out over here at Trovata.

Saved by AI

AI, we’ve established it’s got a lot of buzz, it’s got a lot of money behind it, but how’s it going to impact your lives at the office and is it a real thing now? 

Who better to tell you than the guys at Trovata? We got Brett Turner, we got Joseph Drambarean. I’m going to start with you because you actually have used this stuff. 

Now, tell me about the genesis of when you first heard about this AI thing, when you started playing with it, and how it trickled down and come into market with the product?

It’s a great question. Honestly, this is something that we were laughing about when we first saw ChatGPT come out. And it’s not like ChatGPT hasn’t been around, it’s been in incubation for years now and we’ve seen examples of what it can do for at least a year. And it’s funny that what caught the general population’s attention was an implementation of that technology in the form of a chatbot, because we’ve had chatbots.

We’ve had chatbots-

For a really long time. The Turing test is literally a chatbot. That’s the entire idea.

And the funny thing about that is that it’s the ability to be able to demonstrate the effectiveness of, if you will, a benevolent AI, an AI that can intelligibly respond to any query. 

In the past, chatbots were very limited. You would have to give them a specific context and maybe they’d be really good at a customer service task, or maybe they’d be good at a specific area and you would see it, airlines would use it, banks would use it for customer service, and it’s because if you give it a narrow focus, it could be effective.

The difference is that ChatGPT had this crazy amount of knowledge. It literally learned against the entire backdrop of the internet, which meant that any topic you throw at it, it could be finance, it could be mathematics, it could be philosophy, it could be law, it doesn’t matter. It has pretty much read every textbook that exists about that topic. 

So when we first started to look at it, we were laughing because we thought to ourselves, “There isn’t really anything new that has occurred here.” And the other implication of all of it is that it has not solved the old basic problems that large language models still have, which is being able to do simple arithmetic, being able to infer, in a complex mathematical series of equations, the right answer.

Investing in AI

Let me just stick a pin in that and talk to you, Brett Turner, you’re CEO of this company, you’re in the business of allocation. Ultimately, the decisions you make in terms of where this company is going to focus are going to determine a big chunk of the success. 

How did you see this as something different as opposed to just a flavor of the day long shot in terms of putting Trovata resources towards this?

Well, I still remember too, I was having breakfast. I think I’m talking to Joseph while I’m eating breakfast. Everybody knows, I love breakfast. Most important meal of the day. 

I’m talking to Joseph and he’s like, “Did you hear what just happened?” And he starts telling stories and just the stuff that was the hallucinating event, like at Microsoft, and it would seem like every other day there was something popping up that was just crazy. 

And Joseph all along has been experimenting with this and we’re trying to figure out, there’s aspects that… Trovata’s story is leading up to this, so all of the underpinnings we’re preparing, setting the table for something like this, and so having this as a breakthrough moment and then-

We weren’t new to AI. 

That’s the thing that, we don’t talk about it often enough, but we had been working on AI technology for years at this point. 

It’s easily four years. And the reason why it’s so interesting that ChatGPT has caused this inflection is because you ultimately need to have a foundation in AI technology to take advantage of these types of solutions. 

So there’s this division in the marketplace and tech companies that approach AI. There are companies that can build on top of other technology like ChatGPT, and they might take the abilities of ChatGPT and do something else to automate your life. And there are a lot of examples of that. 

Some of the most fun ones are ones where you can book travel. “ChatGPT, hey, tell me how to have the most amazing vacation in Capri and figure out all the flights that are the cheapest and the best hotels,” and all of that. Amazing.

Which sounds like fancy Yelp to me though.

It’s fancy Yelp.

But in terms of, now it’s not the first product you’ve brought to market, so how is this different for you from that side, from the development side and in terms of creating applications that actually are functional and change the way people do their jobs?

So what we had done is focused a ton on the AI front on the mathematics side, being able to do calculations and infer insights out of data organically in a way that created pretty incredible experiences for our customers. 

Where ChatGPT introduced a breakthrough is by taking its ability to understand any topic under the sun and combining it with our AI and analytics capabilities to create this system where you could ask it anything. 

And then because of the power of our platform, it could respond with accurate mathematical responses that look at any part of your data and provide value. And that’s where we had that aha moment where we saw that and we thought this could be big because it provides almost an extra hand on the team, and that extra hand happens to be an expert in treasury, in accounting, in law, in all of these different topics, and it also has the ability to do accurate calculations. 

So now you have the best of both worlds where you don’t have to worry about the hallucination that we saw in very commonplace when people were playing with ChatGPT, because we’re taking care of that, and instead you can use it as a real leverage resource.

If you’re an executive or if you’re someone in a leadership position and you are caught in a tough moment, maybe in a board meeting, whatever it might be, we truly imagine a reality where you can pull up this technology in Trovata, ask a question, and immediately get the answer to that question in factual form displayed right in front of you. 

That is the breakthrough that we’re excited about and that’s where we think that this technology is going to go. 

We’re just at the beginning. And this is true of every industry. We’re already seeing vast leaps of this technology, whether it’s in the form of chat, whether it’s in the form of diffusion, using photos and video and all of that. 

Imagine what’s going to happen two years from now, three years from now, five years from now. These things will be able to think on their own. You’ll be able to give it a directive and say, “Operate.”

Which is so far past the Pope in a fuzzy coat and other uses for it. 

Now, Brett, in terms of what this does for finance itself, there’s back office functionality. We know what the internet did for every retailer. It went from, we got to be an online retailer, an offline, to now you just buy stuff. And if a store isn’t both on an app in my pocket and at a mall, I consider it rinky-dink, it’s something for everything.

Where’s my one click?

Exactly. So what’s going on, how would you compare this moment with AI, with what Trovata’s doing, and with that prior revolution?

Well, I think the whole vision of Trovata is we’re bringing this automation to these manual or clunky workflows where finance teams need leverage, treasury teams need leverage, accounting teams. And so when you look at what happened, go back 10, 12 years ago, with the rise of cloud and how that massively changed the IT profession, AWS changed so many different things and then Azure and Google Cloud now. 

So when you think they’re just giving leverage to everything, so when you look at IT, instead of patching servers or maintaining your own data center and your own back office closet or you’ve got your own large data center in your corporate offices or down in the basement, all of that had been outsourced by AWS. So instead of managing all of that, that thing that you didn’t really need to manage, and you’re always playing defense, it’s always managing uptime. That was the hugest, the biggest risk and concern.

Now you’re not worrying about that, that’s for the AWS, and so now you could go on offense. So the whole rise of digitization and all these companies and everybody could think about how to then digitize their business and play offense and all the IT folks, really what became more of the network administrator shifted into a cloud architect and all of that started to change a lot. 

So you think of fast-forward to finance now and we always talked about, in those days, the last days of that is always going to be, or what’s going to hit last, is more finance and banking, and it’s finally hitting now. 

Finance finally gets its leverage moment, or its automation moment. It hasn’t hit yet. So despite everything that’s been automated and all of this leverage that most companies get, finance folks are still there. 

If you’re in accounting and you’re having to close the books and do a rapid close, even still, there’s probably times in clothes that you’re not going home at a reasonable hour to have dinner with your family, so that leverage just hasn’t hit.

Because these day-to-day defensive if you will, functions-

Totally.

Are time dominant. It’s regulatory aspects of it, the different models you have to make. Let alone, to your point, before you can even start running hypotheticals, before you can start even asking these questions. It’s like, “I’ve got to close the books, I got to make sure my POS is all lined up. I’ve got to make sure I’m doing my nuts and bolts.” 

Going on the offensive means that you can actually turn your finance into something that’s going to benefit your business as opposed to just the operation that pays the bills and keeps the lights on.

A practical application too is think when COVID hit, everybody’s scrambling and buying toilet paper and all the run of goods in the store, but the finance folks aren’t doing that. They’re hunkered down in a room for the next 48 hours to figure out, are we going to be able to make payroll? 

Is 50% of our revenue going to fall off the map? And for some companies it did. Massive pain. That six months was really scary. And all the finance folks are working overtime because then it’s less about closing the books but, are we going to have books to close? This is more of an operating concern. 

We need to understand our cash flow, we need fast-

And really recently we have those banks closing down, where you’re running hypotheticals that never occurred to you. And it’s not that institutions are unaware that fat tail risk exists as a thing, it’s just that your cost-benefit analysis of what happens if our bank shuts down on Monday?

That’s great.

That was really, you’re not going to assign a staff of five to go figure that out for you. That’s just not a thing that was really likely enough to happen, yet it did.

Yep.

And it catalyzed a lot of what we’re seeing, it seems like, in AI right now.

Yep. Coming from telecom, I think you’ll like this analogy, it’s one of the things that I’ve been thinking about for a while is that within the world of finance, they’ve gone from the Stone Age if you will, in terms of technology, on-prem, using manual processes, using Excel, all of that, straight to AI, straight to the most advanced, most innovative solutions. 

And it reminds me of, if you think of South America or if you think of Africa and how they skipped the whole stage of putting wire down, creating infrastructure, and went straight to wireless, and created 4G, 5G networks right out of the gate and skipped that whole process. 

It’s been really interesting to see how the world of finance, yes, they missed the whole cloud wave for the last few years, but they’re getting to get on board with the best of cloud right now where cloud has weapons that it didn’t have five years ago.

And it’s creating weird moments of leverage that I think are even changing the roles that some of the folks in finance might play and that behavior change has been the most fascinating to see and we get to have a front row seat. 

We actually maintain a community called The Trove here at Trovata where we communicate directly with customers and folks that are part of our community. And seeing their reaction to AI has been fascinating because it’s their first taste of how their roles will change. There are some that are embracing it, some that are worried about it, and some that are in between. 

And having to navigate through that and see that behavior change has been one of the most fascinating pieces of this to me.

AI: The Next Internet?

And is it fair to say that, a lot like the internet, where you have outliers that companies that manage to stay offline and it’s quaint and cute, but essentially this AI is going to become part, it’s going to become a fabric element of the way corporations work now, as opposed to a thing that you can choose to opt in or out, you need to be in and you better figure out how that’s going to happen? Is that fair to say? Is it that extreme?

If you think some of the basics of just understanding, when you start your day, where do we stand? What were some of the numbers from yesterday? Or just all of the usual routine aspects of finance. 

You automate a lot of those things. And there’s processes to automate some of that stuff today, but there’s a lot of these advanced early warning signs, or you take that to an entirely different level to be able to bring up stuff that could be missed easily, because it’s this very rule-based, “I’m only going to report on these handful of things. This is my dashboard, this is what I get.” 

But it’s very limited. 

So when you look at now AI, just putting that to work, now you’re getting alerted to all these other things. So we all see Iron Man and-

I was about to go there, to J.A.R.V.I.S.

Yeah. It is always anticipating, it’s thinking ahead, it’s thinking, it’s thinking for you in some ways, just keeping you at your best. 

And I think that whole aspect of optimization, if you’re thinking about best practice in finance, you’re always trying to do that, but that’s a hard thing because you’re constantly having to reinvent everything that you just did and you thought you dialed in this process and now you’re going to flex it, you’re going to make some improvements. 

And in the days of Amazon, it was this bias for action was a core phrase. Well, bias for action, that becomes a lot easier when you can put AI to work and it’s constantly finding ways to improve at every turn, and we can do that and put that in play. That’s just making your life so much easier.

And takes out so much of the complexity because it has the regulations, you can run all kinds of hypotheticals in finance, but it’s no more sophisticated than what if I sold a hamburger to everyone in China? There’s more to it. You would get ripped off by all the unions and transportation. 

The analogy being in finance, you don’t want to make a bunch of wrong steps, you don’t want to assume a bunch of things.

You can’t be wrong. That’s the thing, it’s always happening. 

We used to joke at Amazon, when you work all these late hours and you have earnings release, and then earnings release is happening and the call is happening and then we’re just sitting back just exhausted, and then we’re joking and say, “Okay, who’s going to get fired this quarter?” 

As in who made that mistake? We’re all going to find out. 

Because you’re working so hard to compile all this information and you’re spending very little time reviewing to make sure that the information is right, and that was the biggest fear. That’s a huge fear in finance. 

So now you’re like, to be able to automate and really focus on more of that time of review and quality control and assurance and just, hey, let’s double click on that and look at how that can really make some big improvements-

That’s the big thing within the business.

This is a user experience change and I think we spend a lot of time talking about the technology of AI, I think more time needs to be spent on the fact that that behavior change is rooted in the fact that your user experience is fundamentally changing. 

You used the analogy of J.A.R.V.I.S. Well, Tony Stark, the reason why we love him as a character is because he seems to operate at a speed of intelligence that is almost at another level. 

Well, a lot of it is because he has accessibility to information and any creative thought that pops into his mind, he can get immediate feedback on whether or not he’s on the right track by bouncing it off of J.A.R.V.I.S. because J.A.R.V.I.S. can then do complex, wild scenario calculations, whatever it might be. 

And Tony, all he has to do is say, “Run me a simulation of me being able to do a space flight in my suit. How long will I last?” Then J.A.R.V.I.S. is like, “No problem, 10 seconds, I’ll figure it out for you.”

In the same way, the world of finance is similar to if you are a writer in Hollywood or whoever it might be, you experience writer’s block. You experience those creative quagmires, if you will. 

And the reason for it is because if you have a thought, “Hey, I wonder what this might be,” the penalty for that thought is, “Well I have to go do the work of putting all this together so I can see if that thought was worth it.”

It stifles creativity.

Completely stifles it.

Yeah, because you think about, “Hey, this is a good idea, but I’m not going to do that because that’s going to take three weeks of work.”

Again, you return to the risk reward and the hypotheticals. What happens if our revenues fall 25% in the case of COVID? What happens if the customers don’t show up at all? Well, that seems crazy. We’re a shopping mall. What do you mean the customers don’t show up? That’s impossible. Turns out it wasn’t.

It’s happening.

And that’s where the breakthrough happens, I think. Because if you get used to that kind of behavior and you start to trust that system, just like Tony can trust J.A.R.V.I.S. flying into space and needs a last second analysis, you start to operate differently, at another level, a different level of speed, a different level of intuition and creativity. And I think that’s where AI, if you can strip the fear of it’s going to take over the world and be ultimately Terminator, you will be at another creative level, because you are operating on the shoulders of a knowledge base that you can possibly store in your own mind, and that’s what’s exciting about it.

And you’re just so outgunned, the people that don’t have this as a tool, who aren’t using it.

Absolutely.

Again, it’s the point of leverage. If someone has that much leverage over, you’re rinky-dink.

So many times too, even as a manager, your team is just constantly tapped, they’re just maxed out. And you have people on your team also that are maybe a couple of the ones that are just rock stars and you’re always giving them the hardest things. And you might come up with an idea and it could be really impactful for the company, but there’s these pause moments you say, “I can’t give it to my team right now because if I do, they’re going to kill me, or they’re just going to spill over. They’re too…”

They’re going to tap out.

They can’t take on any more. That’s constantly happening right now in finance and all that’s doing is at the detriment of the company, the company could benefit from that. 

So I think that just becomes this natural, because you’re overflowing with all this work, just this natural stifling on creativity. 

Another aspect too is that you’re… So many times I go into a board meeting and you spent so much time preparing all the information, you stayed up late at night, you’re not in a mode where you feel like you’re on your A game to present, or you’re going to talk even articulately about some [inaudible 00:24:11].

There’s that balance. The more prepared you are, the more tapped you are presenting.

And you’re so in the weeds, dealing with all this stuff, it’s hard to pull out and talk about something macro when you’ve been dealing with something that’s so micro and I think that aspect is going to change as well. 

So these are very practical things. These are huge things that are going to really tip that whole 80-20 or 90-10, all that compilation, all of the review and creativity or just benefiting the business, creating efficiencies, that whole paradigm is going to be a massive shift. This is all coming. 

I think the biggest barrier, maybe the segue to talk a little bit about, is that, we’re already seeing it a little bit, now that we’ve released it in the product, the first real GenAI tool in finance and treasury, we’ve had companies right away saying, “I need to get permission to use this, because we have some-“

It was the first thing that happened. Security risk would obviously be emphasized. We’ll get into that. I want to leave this segment here. 

Hey, thanks, you guys. Listen, AI, it’s here, it’s right now, and the risk is you don’t weaponize your finance, you just build a little castle and you wait for the barbarians to get here. We’ll keep you in tune at Trovata to avoid that horrible fate.

Permission to Use AI?

AI, it may or may not be the bad guy in your multiplex, no spoilers here folks, but it’s here now and there are some risks with adopting this brand new exciting technology. 

Brett and Joseph are here to discuss. 

Brett, talk to me about the security risk first. As the CEO of this operation, people aren’t that excited to hand over their financial stuff to the internet for really good reasons and leaks in the past. How is this different? And just structurally, let’s start at the real broad brushes of why maybe some of these fears are a little overdone.

Totally warranted. That information has to be protected. And we’re also coming off of this wave of tech and digitization and now the ensuing fraud and cybersecurity and all those concerns that have hit and been in the press and we all know those kind of things. 

So rightly so, everybody’s reeling from that or on edge from all of that, including every boardroom in America. And so all of a sudden now comes this, and it’s like wait a second, that’s even [inaudible 00:26:23].

You’re going to share everything with a chatbot and that’ll make it somehow safer, I get that people are scared. It seems like there are elements that are a little scary.

I think the biggest thing is then how do you apply this in a way that can be safe? How do you ensure that folks can use it where it’s being protected and being [inaudible 00:26:41]? 

And that was one of the biggest things we talked about from the beginning that was part of this, “Hey, we’re going to commission this really cool project but we also know that we have to do it in the right way.” And like I say, all the stuff that we’ve been doing is preparing for that and what’s been cool is Joseph and the team knew just how to do that, which is really cool.

So Joseph, how did you do that? Tell me about how you put safety and made it a part of this application.

Well, we didn’t want to break the experience of the novelty of what you get out of ChatGPT, and what we saw was this clear division. The data that our customers have should never leave their instance. 

That was a rule that we set from day one. But we wanted ChatGPT to be beneficial in a wide range of contexts, and the only way to do that is by being able to communicate with it in a way that is not providing private access to data, but is giving enough context to say, “Hey, you have knowledge of how Trovata works as a system and within the constraints of that knowledge, give me what I need to look up the information to solve this user’s query.”

And that was really where we approached this problem from was that balance of taking advantage of ChatGPT’s natural abilities, which is, it can take any sentence that you throw at it, any paragraph, you could literally write a whole novel and throw it at it, and it’ll be able to take that, condense it, and infer, what do you mean? What are you actually trying to get at here? And then come up with a solution.

So where we got involved with ChatGPT was effectively giving it knowledge of how our database is structured. Not the data within it, but how it’s structured, which gave it a superpower. 

It now knew, “Oh, Trovata has access to transactions, or it has access to balance information, or Trovata knows how to forecast, or it knows how to do analytical things.” And because we taught it that, it would be able to combine that knowledge with the query, the question that the user asked, and say, “All right, now that I know that I can look it up, how would I look it up?” So the last step of all of that was finding a way for it to translate that back to us. And we, again, didn’t want to share any private information across that.

There’s going to be all kinds of information you don’t want on there. You don’t want your competitors to know your payroll, your cost of goods sold. None of it, basically, it’s all your personal info.

So we translated it all into tokens so that it never really knew the language of what it is that you’re communicating back and forth, it just knew that there’s this random string that you’re looking up in a database that is structured in this way, and the final little piece of the puzzle was, 

“Well, we don’t want to ever give it the ability to look anything up in Trovata, so how could we do that? How could we make it such that it only gives us instructions?” And that’s where the math came in, because we didn’t want it to do math, because we had seen-

It’s not good at math and it’s not math you want it doing. Those two reasons are good enough.

So we realize that it’s actually amazing at programming, it’s excellent at producing code, and one of the languages that would be useful within the context of our database would be a query language, a SQL query language. So that was the breakthrough. 

All right, we have context, we have knowledge of the internet, all of the topics under the sun, and now it has the ability to give us instructions in a meaningful way, in SQL. So we put it all together and we created this dynamic interface where you can type anything you want in our chat interface, it sends that in tokenized form up with the context that you need, up to ChatGPT. 

It figures out, “Oh, this is what you mean. I know how to look that up in Trovata,” and then tells Trovata how to look it up. We then look it up on the behalf of the customer and then display it within their Trovata interface. 

That data never leaves Trovata, it’s never given to ChatGPT, and that whole interaction is only one direction. It’s, “Hey ChatGPT, just tell me how to look it up.” They never see anything else from there.

We get accuracy to our customers, which is the huge concern. Again, you can’t be wrong. If you’re wrong, somebody might get fired. Or somebody’s going to rely-

If you’re lucky, someone gets fired. If you can ring-fence it to a guy you can fire, that’s not a bad mistake in the world of finance, on a huge scale.

And the other thing is just reliance, because eventually that information is going to get relied on at the very top level and therefore you can’t be wrong. So that’s always a core tenet. 

The other one is just having this secure translation layer, or this translation fabric that we have. So it’s inherently secure for folks to use and I think we’re already starting to go through that, through the InfoSec process, and we’ve built in security from the very beginning because we’re working with some of the largest companies, we’re working some of the biggest banks, and so we had to do that. 

So once again, we’re already preparing for that. But now to be able to say this is inherently secure so that way you don’t have to worry or you have to do a certain thing a certain way just because you don’t want to trip those concerns, the playing field is already set in a superior way for you to do that.

There’s also a scalability angle to it. We didn’t want to be constrained only to ChatGPT, and that’s the reason for this exchange. If tomorrow, Bard releases as the best experience that you could imagine and a big part of it might be multilingual, because there are clear limitations with regards to the current implementation of ChatGPT. 

Let’s say that I’m in Asia and I want to use this in my local language. Well, good luck. You’re not going to be able to do that. That kind of limitation is not something that we’re used to with regards to our customers because we work internationally, our customers are based all over the world, and so we needed a way to be able to scale our AI infrastructure with any generative AI capability, whether it’s from Amazon, whether it’s OpenAI, whether it’s Google, it doesn’t matter. 

We could be able to take advantage of those solutions, Microsoft being another one, and plug them in as appropriate by region and still have the exact same security model.

So the developers of AI are all operating within basically the same scale, they’ve got the same standards, and that in turn means that Trovata’s not reliant on ChatGPT or anyone else. You can be agnostic to whatever that platform is. And because they all work essentially the same, the idea is you speak to it naturally.

And we’re keeping a close eye on this because at the end of the day, there may be models that emerge that are much better at certain contexts. 

We happen to be using ChatGPT, it’s impressive as it stands, but let’s say that a year from now there is a company that creates the world’s best finance-based LLM. 

Then suddenly we can plug that in and have another breakthrough for our customers, still built on the foundations of our own AI technology. So that’s why that scalability really matters to us.

And we saw that happen in search. Google wasn’t the first in search. You look at Alta Vista, Yahoo, and then Google steamrolled that, those companies, so who knows where this is going to go? 

Clearly OpenAI has a big lead, but even, I think it was this week, Elon is announcing his new play, of course.

xAI.

At the end of the day, your customers win and you as customers win, there’s a certain inception level, you’re a company within a company operating within an industry, all of which is changing at the same time. Trovata’s ability to create this has been enhanced by AI. 

In turn, your ability to help companies and your customers leverage this, which is the huge part, which is so exciting and scalable, that happens either way. That’s just AI itself, it’s not ChatGPT dependent, because all these major, major trillion companies are investing so much money in developing this platform, tells you two things. 

A, you’re a little guy, you don’t want to stand in front of it, and B, you’re going to have a choice. It’s going to be a buyer’s market for this out there because they’re going to give it to you for free because Elon wants to be there just for his personal reasons.

We have no idea where this is going to go, and that’s said in a respectful and humble way. As of right now, we’re conceiving all of this in the form of text. We haven’t even considered it in the form of actual speaking, or in the form of video.

Imagine these types of technologies available while you’re on a Zoom call and it’s just literally looking up all this information and providing it to you in real time as you’re having a natural conversation, maybe in a board meeting, you don’t have to ask anything. 

So this is possible and it’s just a matter of leveraging technologies together to bring it to life, and that’s why we want this posture of flexibility, because we’re discovering this as we go with everyone else. 

The cool thing about our journey though is that it’s applied science, customers are actually using this today, giving us feedback today on how it’s being used. And that’s the most exciting part about it, it’s being able to see real behavior change, the shoulders of this technology.

We referred to you guys as an early stage case study, because so far ChatGPT itself is the case study or billion whatever dollar company, but you’re an actual company that’s helping other companies apply this right now here today. And you’re saying the benefits, the feedback you’re getting from your customers is positive, and you built it in a way that it’s going to provide flexible structural growth opportunities without taking on a lot of risk. 

Is that the gist?

Yeah, we’re already iterating. The other thing is we came out with a release and released it to 

some of our core customers early on. It’s now generally available in the product for all customers and that was about a month ago. And now we’re already iterating where we’re putting it into our forecasting capabilities. So now you can be able to add that and propel that forward. 

We’re just warming up. This is just scratching the surface. But I think just in all the years of just being in startups, for me, there’s always been these moments and venture capital will go nuts and pour all this money into maybe some of the suspected winners or potential winners. And another good example would be internet of things. 

When the IOT thing came out, it was almost like this great solution for… And then they’re finding a problem. Well, do we even really need this? Maybe in 30 years. Or you look at also all the stuff that’s going on with the metaverse and all of that, and so there’s these stops and starts around stuff like, is this really beneficial?

Our goggles are incrementally better, if only VR wasn’t just a horrible experience in general. But this, it’s AI, in contrast to that, it’s something that’s going to be driven by the demand for it. Right now, there’s so much money going into this space, your company better be on board, you better be doing it smart. You guys are in the business of translating and helping other companies stay on top of this and weaponize their balance sheet, in effect.

I think that bringing up the metaverse and all of that is an interesting analogy because Apple just recently announced Vision Pro and it’s in classic Apple fashion, they took an existing idea and just polished it to the nth degree and presented it. But again, what did they do? They focused on experience.

They focused on the details that would give you that anti-nausea effect that you probably are referring to when using VR. Making sure that the screen resolution is so high that you literally feel like you’re looking at real life. Focusing on details about how you interact with this VR experience so that it’s not overwhelming and you can tune it in and out. 

I look at that and I think about how Trovata approaches a lot of the design thinking challenges that we embark on for our customers because at the end of the day, our customers have to love it. 

We’re not just creating technology just so that we can make a press release and say, “Oh, we were the first to AI.” We’re backing it up because at the end of the day, if our customers aren’t using it and they’re not getting any benefit out of it, what is the point?

And the focus on experience, making sure that, for example, the math is accurate. It seems like such an obvious point and I almost feel dumb bringing it up over and over, but you get the math wrong once, one time, and that’s it.

Feed through everything.

They will never look at that experience ever again. And so we had to really focus on, how can we do it? How can we do it in a way where it takes a advantage of all of this technology? 

There will be other challenges, I’m sure, but that maniacal focus on these types of details, whether it’s privacy, whether it’s accuracy of information, whether it’s experiential, finding the best ways to visualize all of this, it’s where our focus is and it’s why we think that we can continue to push the boundaries here, and it’s because we have a vision for how this might actually play out.

Brett Turner, last question to you. In five years, are we still talking about AI or is it so integrated into the way we do business that it just disappears into [inaudible 00:39:40]?

Yeah, it’s so integrated. You look at the early days of the internet, a lot of that stuff, and then even how is the internet being applied? Yeah.

The days of .com companies, now it’s just Amazon.

Yeah. And you look at that first wave and then the second wave has optimized all of that, and the first wave, it was all about speculation and filing an S-1 just after you file your incorporation documents with Delaware. 

It’s like, really? But all of that subsided. Once you got to the second wave, you had big companies, profitable companies, really great businesses leveraging that, that was at the fabric of what they did. I think we’re going to see that really. Everything is moving faster too. 

So I think even how that gets applied, it won’t even necessarily take 10 years to get there. Where you started in the 95, 96, started getting some momentum, you saw a big crash, you saw everything get reset in 2004, 2005. It’s going to take off from there.

I think now that we’re having this bit of a contraction or pause on the economy now, this is, again, probably one of these catalysts, we’re already seeing it pulling us out of that, I think that momentum will continue. I think the adoption of this is going to be pretty fast. 

But we’re seeing that things come out, whether that’s the writers and actors to strike, or Sarah Silverman for suing OpenAI and Meta for copyright infringement. 

All this stuff is going to work itself out, it’s going to get resolved, and then you’re going to see a broader adoption, and doing it in the… All the folks who don’t have the right intentions, or maybe they’re just trying to capitalize on the moment without having anything real and substantial, that stuff, it always works out.

And what makes the moment organic is the breadth of the experience, that it truly has changed society. We’re talking about it from actors to plot devices in movies, to corporations every day to the investing world, which is basically you’re not raising money without AI, at least, is part of something that you’re thinking in terms of the business situation right now.

Yeah, exactly.

It’s exciting stuff. Hey, change is scary, but not changing is fatal. These guys are Trovata. They’ll help you organize it and keep your world in order.