Craig Jeffery on APIs, AI, and Treasury’s Accelerating Tech Evolution

Brett Turner
So here we are on Fintech Corner, another episode. I’m guest host, founder and CEO Brett Turner with Travata. And we have another distinguished guest on the show here. And so excited to introduce Craig Jeffery a strategic treasurer. So lots to talk about, a of cool topics, a lot of buzzy topics. But looking forward to dig in. Maybe first off, just would love to have you introduce yourself and talk about your background.

Craig Jeffery
Yeah, I’m Craig Jeffrey, the managing partner of Strategic Treasure. So we’re a consulting firm, a research firm, and we put out a lot of content. It’s good to be here, Brett.

Brett Turner
Awesome. Well, you’ve seen a lot. I mean, you’ve been in the treasury space for a long time. You were a practitioner once upon a time or earlier in your career as well. So we did a series called Legends of Treasury over the summer. And I think it was just a matter of time to get you on there. We now have you on there. So I would say this is sort of a late entry for that Legends of Treasury. I’d put you in that category. You’ve done a lot. You’ve contributed a tremendous amount to this industry. A lot of people know you, obviously, or maybe everybody knows you. So maybe, you know, talk a little bit about like, being in the space. 25-30 years, what have you seen? Maybe talk a little bit about like, just high level some of the changes and then maybe a story or two that you could say that, you know, definitely would prove out your credentials is that you know, you’ve been in the space for a while.

Craig Jeffery
Sure, so I’ll give an example from more than 30 years ago. So this is really old, and people may not even know what some of the things are. As I worked for an insurance company and we had to pull in bank balances and transactions from multiple banks, hundreds of accounts and many, many tens of thousands of transactions every day. And the way we did it is we used a software called ProCom Plus. We wanted to have it scheduled and there was no such thing as a scheduler at the beginning. So we had a Radio Shack timer that had the seven days of the week. And so it would turn on. So we’d set up a batch file in the computer and have the disks ready. And so at six in the morning, the computer would turn on, the batch file would run, it would dial out with these huge scripts, download all the information, just run, we’d put all the timing sequences in there and download it. So we came in, we would have, you know, several hundred thousand transactions downloaded that we can then process in a database.

And so when you think about that now, it’s like, like, what is this? It’s an old, it was a teletype emulation, which was, which was the most rudimentary way of connecting. And we used a radio shack button with a disk in the drive. Some people may have no idea what that is. And then we got schedulers and wind, and other things that made life so much easier. Yeah. And as you see the progression, and it’s nice to say, let’s get rid of the mess. And let’s make it easier so people can really think about stuff. But that’s one of the nastiest things we had to do. That’s crazy. But it saved us, saved us hours and hours of time.

Brett Turner
This is like a movie or MacGyver putting together, you know, all this crazy contraption, but there’s not high tech. So you gotta go, let’s just go to Radio Shack and get that piece. So literally, if you didn’t have this Radio Shack piece, it sounds like that was like point of, you’re basically alert system or your notification system.

Craig Jeffery
There was a timer that people used to put on their lights, but it went every single day. But we got the fancy 70. One time that eventually wore out and we had to go find another one. Then we got into it just got so much easier as time went on.

Brett Turner
Awesome, awesome. Well, you’ve seen a lot of space, lot in the space, obviously, since those days of Radio Shack. So again, part of starting Strategic Treasurer or kind of getting going. mean, you’ve you’ve seen a lot like maybe walk through just a little bit of the timeline of just things that have kind of happened, you know, over the years. And, you know, here we are, obviously, and we’ll dig into it, you know, AI, all the latest.

Craig Jeffery
So really I started doing a lot of consulting with Wachovia did a lot more consulting did some with a bank before that SunTrust which is Truist, Wachovia, which is now Wells Fargo and It was it was there that we started to see the transition away from installed software to application service ASP software and then eventually became software as a service and so each iteration, things get better by two three four five six or eight times.

So much of what you’re trying to do is the same, but you’re trying to do better, faster, less errors, make it more resilient. And couple that with what people are willing to do. People aren’t willing to, you know, get a Radio Shack device and do this. Right. They don’t want to log into a portal, download stuff. They want to analyze. They want stuff that’s that’s longer term. So that’s probably one of the one of the key areas. The other element is people want systems that are resilient, can expand. And that’s a that’s a part of the system, the technology. I would say that a lot of those types of tasks have fallen by the wayside. They’ve disappeared and been replaced with more analysis, more risk management, more dealing with FX, tighter integration with accounting, with payment security, with coding management. Those are all really, really good things and good developments in the space.

Brett Turner
Okay, if you think about these inflection points, new technology coming into the space, and then the adoption curve. You know, obviously, the treasury space is even within the overall finance org is you have to be risk averse. I mean, that’s it’s part of the job. It’s being a risk manager is a big part of the job. What’s that meant to maybe the adoption curve? When you see these new technologies?

Have you seen those kind of those same trends, early adopters, how it starts to get mass adoption? I think and then one of the maybe the two part question there is that what’s what’s that curve been like? And then also, if you a lot of it’s a it’s a pretty close net community. I mean, there’s so many large corporates, there’s so many treasurer jobs, and these are the thought leaders. And so they’re really networked well together, and they’re talking and, and you’re kind of in that, you know, obviously, in the epicenter of those kinds of discussions and stuff, how closely do those conversations go? And then that option curve sort of speeds up at a different inflection points because they see proof points and then they sort of adopt what they see is working or maybe what’s that been like that might be different than other industries where they might take something and move more quickly and things ramp a lot faster and, know, attack and maybe other spaces.

Craig Jeffery
Yeah, mean, well, tech, that’s a real issue, right? When you have something in and it’s working pretty well and it it takes a certain amount to provide the care and feeding, you don’t necessarily rip it out when you have 15 other significant projects on the table. But there comes a point when you have to get rid of the the bag phone, right? You’ve got to move on to the next level. And so, you if you talk to us about adoption, if we go from the installed solutions, let’s say what was called the Treasury Workstation, you went from the installed solution to ASPs. Once they really, they really started coming out in about 2001 in the vendor provided space. Banks started about a year and a half earlier when they moved off of bringing a disk into companies, installing it, and they thought it was great. And then they had 5,000 customers. when they wanted to do an update, they had to run around to 5,000 customers with disks and then 10,000 customers. And so it just wasn’t scalable. And so this movement, when we went to, let’s just call it the beginnings of software as a service,

When that started, it really started around late 2000, 2001. And by 2004 or 2005, the majority of new systems were coming in in the SaaS environment. It was over half and that was because it was easy. then by, you if you go 10 years later, was like there was almost no, there’s very, few systems that were installed. I mean, they were all multi-tenant, they were all installed. And the same too is, you you talk a lot about cloud native or using the latest, you know, the microservices, small functions, which allow you to make changes at a rapid development, instead of an update every 18 months and instead of an update every two months or one month, now you have, you can remove components, replace, pause that, put in a new component, and you have constant updating. I mean, that’s transformational. So that, we see the development cycle from 18 months to about two months to one month to now continual. And that speed up cycle matters for development. So we see development, you’re probably moving five to eight times faster.

Brett Turner
So if you think of that in this adoption curve, new technologies introduced, obviously coming from tech to kind of help really propel the industry forward, customers forward, maybe the banking side. Obviously, at the end of the day, treasury is an extension of the banks, really managing ultimately your cash and then everything that sort of ensues and builds off that. So the banks are obviously a critical part of that. Connecting the banks, your bank data is a big part of that. How?

The banks even, and maybe this is a segue, and it’s kind of talking about some of these technologies that are moving, and it feels like everything is kind of speeding up, but talk maybe first about the bank’s part in that. I mean, in some ways, the banks are slow to adopt. And it’s not necessarily a pejorative or a slam of the bank. Trovata, we work with banks, it’s big part of our whole thesis is we’re helping the banks get there. They’re not tech companies. And so who’s gonna help kind of bridge that?

Banks are working off mainframe computers. They’re kind of slow to adopt even on the cloud. Even JP Morgan, who would be the most innovative, really just got the green light from Jamie Diamond two, three years ago, even starting using the cloud or AWS. So how is that maybe, that other side of things?

Whether it’s technologies there that are holding up, the risk posture of banks kind of holding it up, how has that impacted that adoption curve? And in some ways, maybe created a little bit of this pent up demand, maybe that we’re starting to see now, but maybe the dynamics of customer demand, there’s a risk aversion on that side. But then also, sometimes you might only be able to adopt this and move things fast along if, as long as the banks are willing or their willingness, maybe, or whether they even if they want to sometimes,

Maybe they’re kind of hamstrung by that. Does that make sense? I know that was a loaded question, the bank maybe involvement in that whole adoption curve process.

Craig Jeffery
Yeah, banks, think if someone hasn’t been in a bank, it’s really hard to understand how many systems they have that they pull together and how many different versions of things they have. You’ve got a DDA system that’s managing this activity. You got a wire platform. You’ve got different payment system platforms. You’ve got posting systems, checks, all these feed together on this really well orchestrated schedule. But it tends to be batch oriented. And as we move to a streaming environment. You can say not everything has to be instant. Nothing’s moving slower. Everything’s moving towards streaming. But you can’t make the jump for everything all at once when you have, you know, say a trillion-dollar bank, they’ve got multiple, multiple areas feeding multiple, multiple systems at all stages of upgrades and development. So this move to streaming, you know, like more like the API.

Brett Turner
Streaming more and more, like transaction to transaction.

Craig Jeffery
Yeah, as opposed to we process this at night or we process this at two times a day. So as things happen, they occur. So there is a move towards faster the API world streaming. You see that in certain companies as well. Not everything has to be immediate. Nothing’s moving slower. More, there’s many, many use cases for immediate or quick. How about you look at accounting close. Used to be two, three weeks. Now, it’s like two days later, they want to close. That’s accounting. The treasury wants things faster. Banks want faster so banks are dealing with that. The ones that can handle it, do banks handle it? The trillion dollar banks, they’ve got huge pockets and they’re working on it. They’re making changes themselves. They’re bringing in other vendors to provide API gateways to their backend system, the self-service API kits that they put out there. So they have the resource to make it work. And then you get down to the smallest banks.

The smallest banks are almost 100 % dependent upon their core banking provider because they’re going to use what those things are. So if they want to connect APIs going into other systems, they’re fully dependent. Then there’s a mid-range where…

They may be in a larger system that may take a little longer time. There’s at least a tale of three worlds, I think, at least in the bank. But to your point, the thoughtful banks say, I want to be able to deal and work with my customers where they are, wherever they want to work with. So this embedded banking or embedded treasury.

Brett Turner
So maybe this is a good segue to talk about. So digital transformation, maybe let’s kind of go through the pecking order of this timeline and now things are accelerating. You look at APIs, and we’ll talk, kind of a good segue off the banks, because if you think of if demand is starting to rise and started to kind of go, you know, that noise gathering more of a crescendo and the banks aren’t able to meet it. I mean, even like you mentioned, community banks, think that traditionally has been the case. But even like even really big banks now, the demand is so high and the agility and requirement to move really, really fast on so many different industries. It’s even too much for the banks to kind of invest in these kinds of things. They don’t develop these things fast enough. That’s not necessarily in their DNA to do.

So here they have embraced APIs, which took a long time. The tech world had been using them to kind run everything for 10 years. That’s the world I kind of come out of. But here all of sudden, banks now start to open up. All of a sudden you now have balances, transactions available. It sort of opens up this possibility of coming in embracing that and develop it maybe in a little bit of a different way. I know data’s been around via Swift and files, but it’s sort of an IT hop or it’s sort of a pacifile SFTP.

And there’s not a lot of quality assurance in that layer. It’s like, did we get the data? I don’t know, we threw it over the wall, hope you got it. But now you’ve got APIs as now what could be a building block. So maybe talk a little bit about when APIs hit, where is that maybe today? Has that sort of, maybe the good, bad, and the ugly on maybe APIs? Because now they’ve actually been here for, you know.

Craig Jeffery
They’ve been here, they’ve been used, the adoption is sky. If you can connect and the bank has worked out the kinks, right, when the bank first starts, on, the first comes on, they don’t have the volume scaled. And so it’s like, hey, this has bended, right? Just like every new tech, there’s a little bit of hiccup, but it’s a very short period of time where there’s hiccups, then they solve those issues. And so it’s very, very quick. know, banks are, most banks, I think, that have a mature view of connectivity of data are…

We’ll charge you for data. You want to connect via Swift? You want to do SFTP? You want to do APIs? We’ll sell you the data. We’ll connect with the way you want to connect because that’s how you want to consume it. And there’s going to be a transition. Everything’s not going to change in one day. And so the smart ones are like, we want to support the leaders, but also support the back end until it gets too old to support. So I think those banks, those institutions are making the right kinds of decisions and then saying,

These people are connecting through their ERP, through their payment hub, through another system, an admin system. We’ll make APIs available so you can transact with us wherever you want to. Five core systems, a treasury management system, a payment hub, any of those things you can engage. I mean, that makes the most sense. It’s like, you want to do business this way? We’ll do it.

Brett Turner
Yeah, well you see even too, even the, and here you have something that’s, you have an adoption curve with APIs, but even if there’s technology, you know, a workstation that’s been around for 20, 30 years, and the data model isn’t really apt to handle something maybe that’s unstructured data coming through the API, like even what’s maybe your view, of what’s been sort of the intersection of how to take in something, a new way of to deliver and then having to contend with that, having to adjust to that, what’s that been like over the last few years with some of the technologies you’ve delt with.

Craig Jeffery
Yeah, the last few years in over time. So the formats that used to exist when people passed data were all compressed because you had these little ProCom modems.

Is that pick up on the camera? That’s really, that’s tricky. Yeah, well, I don’t know what that is, so these data costs a fortune to send. So they compressed everything, they put commas. So what what the data was, was defined by where it was located, which is great if data costs you a million dollars to send. Now that it doesn’t, we want data that’s richer and you want to be able to change it. And so now if you have an API or you’re you know, ISO 20022,

You’re sticking more and more information into that system. Now the receiving system doesn’t a bend because position seven now says, you know, some has some other tag in it. Now you can use it or not use it. So that is, that’s how we connect, how rich the data is, is completely transformed what can be done and how rapidly things can develop.

Brett Turner
So that’s a segue to sort of the next thing that’s kind of okay, you’ve got API’s, there’s this riptide that’s happening now. All the existing treasury technologies are used to getting data a certain way. You’ve got standards, a CAMT.053 and yet now you got API from a certain bank who isn’t adhering to that standard; they’re just pushing through. It feels like willy nilly all this data. Hey, it’s rich, it’s great. The customers like, give me all the data, we can, you know, use it for BI capabilities or we can innovate on top. That’s kind of been Trovata story, and we love that. We can handle it. We’re a newer platform, but a big part of our journey has been, you know, because we’re cloud native, we can handle that. So maybe how has more of cloud native solutions and technologies with now the hyperscalers, now it’s becoming a lot more commoditized in terms of all the different providers even having them. How has that resonated? What maybe has been that journey or that realization to say, okay, now we have something that could be really cool to embrace. But now we’ve got to also make some upgrades here to be able to take advantage of it. What maybe has been the importance of that and how has that resonated with customers out there?

Craig Jeffery
Yeah, and we look at as a consulting firm and that does a lot on the technology front. We look at it from a. When do you need to upgrade? Not only what should your which is your future state need to look like, but when do you need to make the jump? And so part of that is your whole I’ll just call your conceptual tech stack data connectivity systems and then we push reporting analytics together. So you’ve got it. You got multiple questions embedded in what you’re saying. If we look at it from the stack, it’s data, how we connect.

Those have to be thought through again. If you thought through that five years ago, you need to rethink through that again. We’d thinking through that with people to say, what do you need to do so that every time you make a change, every time you make an upgrade, it’s when it makes sense, it moves on to the newest and the latest platform, not to a version ago, you’re not putting something in there that has an end in life in view. So I don’t know if I answered your full question, at least it started, I think.

Brett Turner
Well, I think then you look at, okay, you’ve got this data that’s really interesting and there are some customers that say, want all, you know, they might be connecting to 50 banks. I want them to all be APIs for all those banks. It’s like all those banks don’t even have APIs. Even though they’ve been around for a while, not all banks have even adopted and kept up in that. So what do you do then? You got the, again, demand outpacing the bank’s ability to open up and provide that. Even the banks that do have the APIs, you’ve got a different way of data coming that’s richer, can be real time and all that. Can you take advantage of it? Okay, do we have cloud native technologies to have that agility to be able to take it? So that’s been a thing. So then all of sudden, so then there’s been this mode where it’s say, okay, we don’t necessarily have to adopt that yet because the need to be able to do the treasury work requirements, you know, are trumping that so we can kind of be okay. But then all of a sudden, AI is coming along. And now it’s coming through every boardroom.

It’s coming top down. It’s coming through every CEO. It’s coming through CFOs. In some ways, massive wave is just now if you’re hitting other areas of tech, they’re kind of embracing it. And even they’re a little bit flat footed in how this has happened so fast. But now you’ve got, in some ways, a data or technology readiness thing that isn’t quite prepared for this. And now it feels like a tidal wave that’s just coming in. So what maybe has been, you know, maybe that initial, what you’re hearing and seeing, like this initial industry response, seems like customers are excited about it or maybe cautiously optimistic, but it’s generally like, okay, we know it’s coming, there’s acceptance.

You know, how close do you feel like is it here and now? Where do you feel like now this comes because it now it feels like already or not it’s coming and there’s now a lot of pressure from the top. That’s which is a whole different way or organizationally in a way maybe that wasn’t the force in the mechanism for these other things. Maybe talk a little bit about like AI and what that means and how that’s coming.

Craig Jeffery
Yeah, you look at AI, APIs, even if you go back to business intelligence, the ability to self serve and instantly determine things. All those are like three, three rivers merging with great, with a great and increasing speed. So there’s a lot of turbulence. How does that all look? If you’re trying to do a, if you’re trying to do some kind of query or learn something, do you do it with an AI tool that allows you to flip everything instantly and learn and iterate? You do it with a open natural language question query to get something and so some of those like how do we how do we pull those together so that we can meet the final need that’s where I don’t know if anybody knows all the answers to those because there’s so many things that are developing so rapidly there’s good ways to do it in both cases but to your question about AI, is real and significant. People are using it. Primarily, Treasury people are using it within systems like Trovata, like other platforms. It comes baked in. They built the rails for you for protection, like all that’s built in. That’s where that’s happening mostly. Other Treasury groups are, I’ll say, experimenting or doing small use cases within their groups on developing reports for risk, trying to determine what their exposure is across their database or the ERP system. They’re writing things to pull out and extract data that helps them make good decisions. So it’s real. You probably read the MIT report, like how many things fail on the AI side, but the ability to learn is making a significant impact. So people are using it, and it will scale very rapidly. Just seeing how familiar people are with the AI that that generates or creates the AI that detects patterns or extrapolates and the AI that does things like agenetic AI. Yeah, those are we’re gonna see so much transformation in a much more compressed period of time. Yeah, you know.

Brett Turner
It’s crazy how fast it’s moving, and the other thing that is so, the kind of the last piece now it’s coming, and it feels like AI is moving so fast. It’s gaining this momentum faster than maybe any any other technology even the stuff we’re talking about now. It’s I think this forcing that’s a mechanism that’s even accelerating faster, and then you have stable coins. So now stable coin,s which you look at the circle IPO that happens.

It’s 4 months old, one of the biggest IPOs in 20, 25 years, goes up eight, nine, 10 times in the first day of trading, which is crazy. And I know it’s kind of pared back since, but still, it just was a monumental thing to say, no, this is real. This is what people are really embracing. And now, like, this is something that is now kind of forcing the banks to reckon with because now it has an aspect of almost like a new rail, a new way of, you know, how is this put? So now the banks are sort of forced now kind of, again, ready or not, here we come, accelerating even things on the banking side. What are you seeing and hearing about stablecoins?

Craig Jeffery
Well, that’s good. You said sort of like a new rail. I mean, it is a new rail. It’s a new way of doing things. It’s similar and things are significantly different. know, variable rate currencies, I think there’s essentially no use in treasury. But stablecoin, payment stablecoins, you know, the Genius Act that provides a framework which will be in process and probably within a year, provides a really good framework for having payment stablecoin so we can make payments on the blockchain easily. There’s other digital assets.

Brett Turner
What do you think, besides maybe payments, of course, cross-border to others, but what about just being able to settle inter-company transactions? Is that the killer app?

Craig Jeffery
Well why do you need to go through a bank or you don’t need to move money like you never needed people had netting because why did they want to settle those transactions right it doesn’t make sense.

Brett Turner
Well you still need keeps arms length, comply with transfer pricing agreement, those kind of things.

Craig Jeffery
I mean, that’s not the settlement though, right? Yeah. Yeah. So it’s so you’re I mean, you’re exactly right. It’s. You still have to have the right framework and follow the rules. Yeah. The settlement mechanism is why do you need that many transfers? So you net it down and then so you do less, less effects, less trade, less.

Brett Turner
Do even need to do like netting? Again, it’s kind of like batch. You’re doing batch processing for the because it’s a hard thing. Let’s do that. But now, if stablecoins are free and you can treat them, almost put those in the hands through a treasury solution to essentially do your settlement on a line-by-line or transaction-by-transaction basis. And it’s free. And it’s then you’re kind of doing your own ledger transfers, like you even need to wait. Do you even need a netting solution?

Craig Jeffery
Well you you may want you know something that the timing is monthly or quarterly or so the factor that because of of cash flows are things that are systemic but you could like you said you could go as they occur

Craig Jeffery
That’s more towards streaming, You’re up to this, everything. So you have the ability to stream or the ability to go instantaneous. You don’t always have the need. You may not move there, but now you have the capability. And so where the use case makes sense, you’re exactly right. That’s perfect. But just because you can go faster doesn’t mean you always have to. But nothing is moving slower. There’s nothing I’m aware of that moves slower. It’s always faster.

Brett Turner
So if you look, so maybe just a quick, know we got a minute or two, but maybe one last question. So, you it’s out there. We’ve talked a little bit about it. We announced an acquisition we did in the summer of FSC, and kind of having full TMS capabilities on Trovata now. I think one of the things in terms of we focused on this data story, trove of data, Trovata, focusing on data automation of workflows on that data.

AI is obviously a huge part of, you can’t really do that if you’re not data ready. So those are big parts in our story. But now you have with one of the things maybe moving up market to really large companies is we haven’t had some of the key checks and boxes on feature functionality when it comes to TMS. Now we have that. So I would say it’s like here we are, know, world series happening. We just witnessed Shohei Otani. don’t know if this is rest of it, but here is the first, saying he’s the you know could be the the greatest baseball player ever almost babe Ruth like because he goes out and does one of the most athletic feats of all time he pitches six scoreless innings and then he hits three home runs in the same game like how does that happen so I look at that and say hey Trovata now is like a best of both worlds like you not only get all the modern things around data AI but now we have you know this this dual component we now we have all the that’s checks and boxes on treasury functionality. I would love to like your thoughts around maybe what this would mean, what this could mean for the industry, know, accelerating around putting all these things together, best of both worlds. You know, does that mean a lot of the things that now this acceleration of demand, all these technologies where people want to go, can it really meet and satisfy a lot of that?

Craig Jeffery
Clearly, when people are looking at when I’m moving technology, when do I upgrade, when do I make a move, do I not have something, do I need it? You want to move on to newer technology that meets a significant subset of your needs. The ones that are, the newer the technology, the faster development can occur. And the second factor on that is, is the company that’s shepherding that or caring for that, are they pouring into it to stay current?

Because every technology will become obsolete. It’s hard to understand, but everything will become obsolete. APIs will be too slow. I don’t know what’s going to replace them, but if we move towards that streaming, those things are going to have to take place. So part of the decision on the tech is what is going to generate the most value in the near term and over time. What will increase in value over time? And that’s part of a strategy decision. It’s part of a tech decision. And it’s part of a, what’s the network you’re connecting into how big is that network what can it do for you all those factor into the the decisions but you have to approach that too from what does our what is our tech stack need to look like how are we meeting our needs and so those are – I love those conversations because stuff is changing quickly and you have to stay up on it and when you’re making investments no matter how quickly the technology changes you got to think it’s gonna be at least a decade you’re gonna get it up and running optimize it new tech comes then you have to replace pieces of it so

Brett Turner
Yeah, yeah. Well, good. I appreciate the conversation. It’s been great as always. Thanks for being on Fintech Corner.

Craig Jeffery
Yeah, my pleasure.