The New Intersection of Technology and Treasury – Fintech Hotseat Panel Discussion – AFP 2023
The rapid acceleration of new technology adoption significantly impacts the daily operations of providers and treasury professionals across multiple domains. In today’s podcast episode, we’ll hear from Craig Jeffery and our 3 panelists in an annual Fintech Hotseat panel discussion, this time covering the evolving landscape of treasury technology.
Craig Jeffery, Managing Partner of Strategic Treasurer,
Ron Chakravarti, Managing Director at Citi Bank
Kristin Robertson, Senior Sales Executive at Finastra
Jon Paquette, Executive Vice President of Solutions and Product Strategy at TIS
Explore nine years of Fintech Hotseats on our website!
Subscribe to the Treasury Update Podcast on your favorite app!
Episode Transcription - Episode #282 - The New Intersection of Technology and Treasury - Fintech Hotseat Panel Discussion – AFP 2023 transcript
Welcome to the Treasury Update Podcast presented by Strategic Treasurer, your source for interesting treasury news, analysis, and insights in your car, at the gym or wherever you decide to tune in.
Jonathan Jeffery 00:18
The adoption of new technology is happening at a rapidly accelerating pace. And there are a number of areas where this intersects with our daily activities as providers and treasury professionals. In today’s podcast episode, we’ll be listening in on our yearly FinTech Hotseat panel discussion on the future of treasury technology. Our speakers will be sharing insights on API functionality, embedded banking, interoperability, and more. This panel is hosted by Craig Jeffery, Managing Partner of Strategic Treasurer. And our three technology experts are Ron Chakravarti, Managing Director at Citi Bank, Kristin Robertson, Senior Sales Executive at Finastra. And Jon Paquette, Executive Vice President of Solutions and Product Strategy at TIS. And with that, I’ll turn the show over to Craig.
Craig Jeffery 01:08
Our first topic is technology changes and challenges is about API’s embedded banking, and ecosystems. So think about how everything inter operates. I’ll start with Ron, and then we’ll go to Kristen, and then across to TIS. So I’ll throw a few questions out there and let you get started. Feel free to agree or disagree, if you agree, be quick about if you disagree, explain why. What are the biggest shifts or changes in technology over the last few years? And also someone could give a definition of embedded banking at some point during the discussion? And how will we and where will we transact and trade? Will that always be in a bank portal in a technology system? How will that work? And how will that change? And then I’ve got some additional follow up. So Ron, if you would start us off? That would be excellent.
Ron Chakravarti 01:55
Sure. I’ll take I’ll take your first question. You know, what, what are the biggest changes, I think to what I describe as modularity. shift away from monolithic systems into something, you know, maybe 20 years ago, we’re doing the same but best to what a best of breed systems and then tying them together. So modularity, the other one has yet to play out and really have an effect. But I think it’s going to become increasingly important, which is a real focus on data, data strategy, data initiatives, getting the data lake together, and then figuring out what we’re going to do with it. And by that by we mean on the on the bank side, as well as in the corporate side and our corporate clients.
Craig Jeffery 02:33
So interoperability and data. Kristin, what’s your take on that?
Kristin Robertson 02:37
Well, I think with the change, frankly, to adopting ISO really, as a global standard has forced a lot of that change. Because the historically here in the States, a lot of things were managed to old mainframe technology when it comes to the amount of data that flows through. So that’s been really critical, I think, to a lot of the adoption and change in the speed that it’s occurring, especially with the Fed going to ISO in March of 2025. But when you look at you know, from how will that people operate from a look and feel perspective, you mentioned in embedded, and I think that what you will see going forward is you had obviously the technology change for standards, but more and more folks in the business space, are expecting a similar experience that they get in retail. So if you look at like a Venmo experience, you don’t need to know bank account details. It’s frictionless. It’s easy, it’s electronic. And I think you’ll see more and more of an adoption there where things are going to be embedded so that from the end commercial clients experience, it will be seamless and frictionless going forward.
Craig Jeffery 03:38
So is embedded more about disappearing the activities that you think about more than just the tech or?
Kristin Robertson 03:45
Exactly. So if you think about it, when it comes to making a payment having it be a simple, seamless process that you click through on a mobile app, when it comes to submitting invoices b2b, the same requesting for payment, where you actually are able to send invoice data, a request along with an immediate payment transaction request, that will end up ultimately being frictionless. But there’s lots of puzzle pieces to assemble to get to that point, but via microservice as is and being able to and plug that right in. So you’re not logging into a treasury workstation, per se, or a bank or portal. It actually is happening from an end user perspective behind the scenes.
Craig Jeffery 04:21
So friction less or less friction. And John, what any anything to add on the the ISO front or the embedded banking side?
Jon Paquette 04:29
Yeah, I think I agree with most of the points that were made so far. But you know, I think there’s a fundamental shift going on. And what Treasury’s really expecting from technology these days, if you look back to the original TMS is it was really about automation and reporting. That’s what people wanted to get out of those solutions, right? Fast forward to nowadays and it’s more about its insights than ultimately getting to actions right. It’s people want to use these technologies to ultimately be better at their jobs be better cash managers, and that’s why you’re seeing corresponding a lot of interest in some of the technologies that are emerging like API’s. Why do people want API’s because they want to get data faster, so they can make decisions faster? Why are people interested in AI because they want, you know, software that’s going to help them extract out the insights and ultimately qualify a decision much quicker. Right. So I think that’s the overall trend I’m seeing in the market. I think, embedded five, I call it embedded finance and banking. But I think it’s a huge part of that, right? Because it removes all the friction from the process. And ultimately, when you get from that insight to that action, you can complete that action natively within the application that you’re already working within. Right. So concrete, so totally frictionless process I think corporate treasurers are really looking for.
Craig Jeffery 05:30
So shifting to new technology always sounds awesome. But there’s this issue of tech debt, right? You have, you have systems that work. And when you have to move, you can’t necessarily move everything all at once. Can you? How does a company and how does a bank, how to fintechs deal with this, this tech debt? You know, and the idea that if you’re a bank, for example, you have to support companies on a wide range of technological capabilities. You can’t just say, everybody has to move to, you know, XML tomorrow, there’s this there’s this transition period, who wants to talk about tech debt? And how we should think about moving to the newer tech?
Jon Paquette 06:12
I could take it Yeah, yeah, for sure. So I mean, I think the the promise, I can give the software providers probably viewpoint on that. I mean, you need to be flexible, I think to be able to support a variety of different protocols. But you also have to be flexible enough to allow corporates to move to new technologies as they’re ready to. So I think it’s one clear trend we’ve seen is that open banking API’s, everybody’s really interested in them. But no, banks are really providing a consistent experience for corporates, they all have different levels of capabilities, right. And it’s not right for every single corporate to move to an API, it might make more sense for their business to stay on something like a host to host connection now. But they don’t want to sort of curb their their their maturity in those technologies and say, hey, now that I’ve picked a host to host connection, I can’t move to API’s in the future. It’s up to the providers to allow them that flexibility to do that. And I think also make that a seamless experience. So while you’re managing sort of a variety of protocols in this environment, where I guess everybody’s sorting out this tech debt, as you kind of refer to it, that’s right. So that, that, you know, you still have the experience, you’re still getting the value from the solutions that you’re really looking for on a day to day basis.
Ron Chakravarti 07:13
I’ll answer it, as you’ve heard from, from the from the customer side, I would say most companies of any size, also have the same issue. I mean, I’m talking particularly about the Treasury team, there’s a tech debt, there’s a tech debt that’s that exists, because you’ve got obviously what the firm has got in terms of his ERP or infrastructure. If you’ve got a TMS It’s probably some years old, larger, it is more likely aborted some years ago. And it’s very hard to make change, it’s very hard to go back. If it works, you know, if it ain’t broke, don’t fix it is sort of the adage. But increasingly, I think that this maybe this is sort of a link on to the earlier point about people are thinking about data. I think the other thing that companies I find, particularly, the larger the company, the more the propensity is to think about what is going to be our technology strategy going forward. You know, so the company is going at a certain pace, we see what banks are doing, we see what’s happening in finance, generally. And the group treasurer is saying, Okay, we can’t just keep upgrading our TMS, you know, to the next version, we can’t just sort of put in another couple of different systems that do interesting things, we’ve got to think about the tax strategy for Treasury, in particular, and go back and make the business case to the firm as to where we need to be in the context of what the company is doing in technology. But to have a very clarity of where we want to end up being. And I think the, the Italian, we have, we’re on a server called city treasurer diagnostics, we’ve been running for about 10 years. And the bias is towards tends to be towards larger companies, which is cities client base. But 40% of companies in the last couple of years, if I average it out, have said we have got a treasury technology strategy, either in place, or we’re actively working on figuring that out, not just the usual upgrade, not just do we need to do this or that. But we’re actively working and figuring out our because we read and when we go back and ask the question is because that’s the only way we can go back to the firm and get the resources and the budget for what is a specialized function in the finance organization in the context of the vast business of the company.
Kristin Robertson 09:24
Yeah, earlier on to that as well. I think that it’s not an easy question to answer because a lot of larger corporates in particular, ended up customizing historical solutions to really suit their unique business needs. So I think a lot of ways they’d like to get out of the technological debt that they have, but analyzing, you know, what’s available either through their banking partnerships that they have for their commercial cash systems, as well as the number of banking relationships that they have. I think you’ll see going forward and analysis where how do I keep what’s specific to my culture to my organization while it does opting something that’s truly cloud native, whether it’s the banking partner, because there there’s a lot of rationalization in the banking space going on for corporates today, or I think you’ll see more and more small businesses aligning with the banks to provide solutions for for Treasury going forward. And then the Treasury workstation disruptors that are truly multi tenant SAS. That’s where a lot of folks want to go. But figuring out the way to plug that into that, while still having what matters most to them from a customization perspective, is the big challenge when it comes to technological debt.
Craig Jeffery 10:31
Excellent. And I know we’ll talk about some of those terms like cloud native and what that means in a section coming up shortly. But let’s move on to artificial intelligence and machine learning. Many of you have played with chat GPT. Perhaps you’ve used Bard or other AI systems, you’ve asked inane questions or maybe helpful questions, I’ve always found it to be, you know, as I started it, it’s like, Hey, this is like a really good intern that does things in like 30 seconds, and does a better job than interns, no offense to the interns. But you do have to check it, it gives you bad information. And you have to poke it again and again, to try to get right information to correct it. I remember asking about reserve requirements. It was like I know, it was 15%. And I said it wasn’t 15% i Oh, sorry, you know, looks up other data, right? It gives a quick answer just like a person would. But you can grow to depend on it. I wanted to talk about that and hear what your thoughts are. And maybe we’ll start with John on this one, what’s happening now and what’s coming soon and Treasury. But maybe as a backdrop to that, what are people doing and using? After we have that that discussion? I want to hear how that’s going to impact people and staff? What is changing? And is how dramatic that will be. So what’s what’s what’s happening now? And how will that change?
Jon Paquette 11:54
Yeah, I mean, I think AI is changing the way people are working in general. It really is. I mean, I think it’s a real deal technology out there, that’s adding a lot of value in the market already. And I think the best applications of it exactly what you kind of alluded to Craig, which is productivity, quicker decision making, and just sort of getting to that answer as soon as humanly possible through using AI solutions. So within our own business, not even like within our software solutions, but just within our own business. We’re using AI for code development now. So to write code, to complete code for us to translate codes to different languages to extract code from natural human language. And just the productivity increases, we get there in terms of developing our product are just absolutely astronomical. We also use it to proactively identify data that might fall under certain regulations like GDPR, or personal identifiable information, right through just enormous datasets using AI to really sift through and be able to pick these things out. And these are things that really save our business a lot of time on a day to day basis and make us be able to be a lot more efficient in the way that we operate. And, you know, I guess move into, like how it sort of goes into treasury, I don’t think it’s quite there yet. It’s not providing that same level of value that it is in other areas where people have adopted and more naturally, there are companies that I mean, there are some providers out there that have incorporated in like chat GPT, or like, you know, chatbot AI technology, right for that purpose to be able to sort of sift through your data quickly. And then because it’s an easy thing that’s sort of licensed, incorporate into your software as well. And then there’s a lot of confusion around AI, I think in the market as well, where people sort of think that certain technologies are AI that are more like sort of plugging data gaps and things like that, for companies like we come across this all the time, we have actually a functionality in our product where we can sift through based on like invoice level data in our forecasting product, and look at when a customer is scheduled to pay based on payment terms, and when they actually pay based on their historical payment patterns. And people look at that, and they say, Hey, that’s some really great AI you guys out there. And we say it’s actually not a it’s just sort of an algorithm that we wrote. But you know, we’re, we’re glad you guys like it, you know what I mean? So there’s a lot of confusion in the market about that term as well. But I think that when you really see the best applications of it, it will be really geared towards helping people sift through those massive data sets and qualify and make decisions and ultimately make cash management actions much much quicker to the use of these technologies.
Craig Jeffery 14:07
Kristin, you want to jump in on AI and machine learning uses and outlook? Yeah, definitely.
Kristin Robertson 14:13
finastra has also adopted our own AI technology that was we’re using in house for development and other areas as well. And it’s from a FinTech perspective, I think you’re going to see more and more of that. But you raise a really good point, I think they call it hallucinations when it comes to the AI machine learning side where if the data is not good, or the data is biased, it’s only as good as the data that you’re actually leveraging whether you’re using a closed loop, AI solution versus chat GPT everything’s public. So you need to be super aware of, you know, where is your data going in from a FinTech perspective, or the fintechs that you’re partnering with, whether it’s Treasury workstation, you know, the payment providers that the banks use, are they allowing for a He obviously, pool of data to be leveraged not only to see what your own predictions are within your firm, but how do you compare to others of similar size? So for example, if you’re looking at something like what’s my DSO and what where am I going from a prediction perspective? How do I compare with other corporates of a similar size so that you can then have even more meaningful data? I think that’s where we’re going. But data will be king. And being able to actually leverage cross functional data and multi tenant SaaS solutions is going to be really, really important to bringing that beyond things like in line fraud screening, where it’s very, very much used on a broad basis today, because it’s not as sensitive.
Ron Chakravarti 15:39
So if I, if I think of it in terms of and it’s implied in and what’s been said, Is there I think of it in terms of the predictive versus generative AI, predictive models are obviously not new. Most companies are built cash forecasts have do some sort of multivariate regression and work it in and as John was saying, you know, companies for years have got, you know, deterministic logic based processes with the pattern matching, but perhaps some predictive AI, as long as the What the It’s not a black box, I do see everyone starting to test it and stop because the tools have become easier. And again, if you look at Cash Forecasting is sort of the the most classic use case of it, you know, companies and organizations becoming more comfortable with that. Right. So that’s predictive AI. But that’s sort of a relatively narrow, that he says currently envisaged, obviously, there’s been a lot of excitement because of generative AI, right. And the thing with generative AI is that if it’s trained on sort of the data at large, then it picks up effectively, it’s still a pattern recognition engine, really, but it picks up bad correlations, it does it sound hallucinations, which is really false correlations. I think that for generative AI, two things, the power is going to come as the tools become easier to use, and companies and organizations bring it inside their firewall. So it’s trained on data that they have. And as the companies and organizations also improve that data lake and clean up the data. So if you bring it in, and you’ve got a good clean data warehouse and a data lake, then you’re really going to be able to put it to power. But in terms of the I would think of that as if Cash Forecasting is the classic use case of predictive AI. The classic use case of generative AI is this. So this was a real conversation with a group treasurer of a footsie 100 company at dinner with a couple of months ago, right. So I have 100 million Ozzy is sitting in Australia. And all I have time for is to ask now let’s take a look at the cash forecast and decide whether there’s any sort of big need but otherwise is it going to be a time deposit with the bank is going to be an awesome money market fund. And that’s pretty much it all we have time for right there’s a big company has 100 million, it’s not that much for them. But if I have been if I really had but I’d really like to ask the analysts to do is go take a look at that go ticket, the intercompany loan book in my in house bank and see what’s maturing and whether I need to be doing something. And I should really swap it into Euros and do something with it, or whether because of the capex that we’ve got going on in the US blah, blah, blah, blah, blah, I should really be looking for a term loan to that I’d be looking have been looking at a whole bunch of scenarios. But the analysis is in the pot, just getting that data together is what’s going to take you know nine tenths of the time, and then there’s going to be 10% and generative AI, if you bring it in, and if you’ve got good data, you go and be able to query that data to bring it all together. So the human still does the final analysis and decision. But that power of bringing the data together I think is where the real power is, there’s some ways to go because you again, you gotta bring it inside, you’ve got to have the clean data, etc. But I think there’s a there’s some real power to be unleashed in productivity, that that’s come from that, again, for corporate treasuries. And I completely agree with the use cases for providers, whether it’s a FinTech or a bank, in terms of coding in terms of customer service representatives, being equipped with scripts, all of that absolutely been the clients that that’s what I see.
Craig Jeffery 19:00
Excellent. So so just a follow up question to anyone on the hot seat, you have the use of AI and machine learning, either running through a vendor system or running through a bank system. And Ron, you also mentioned about, you know, loading data into a data lake and so the company’s running, you know, AI, generative AI or otherwise off of their, their system, how is this going to progress through companies? Is it going to be, it’s clearly going to start using vendor systems and bank systems and move? How is this going to progress, though? How should we think about that?
Jon Paquette 19:31
Everybody’s looking at me, so maybe I’ll take this one. But, um, you mean in terms of how just organizations will store data long term, how they’ll sort of manage these strategies?
Craig Jeffery 19:39
And how will it move out of vendor provided functionality for AI for pattern detection, learning into companies applying it across their vendors and their banks, you know, into the data? Like, how how long will that take? What will be necessary to make that? I know that’s a hard question, but yeah, I want to know.
Jon Paquette 19:57
I think it’s a good point because yeah, the whole point Under this system is to break down these barriers, these data silos between different departments in the organization bring all this data together and make comprehensive, you know, good good finance decisions within a terminus become real popular, the Office of the CFO, right across all these different applications, but um, I think that the trick is the data cleansing that Ron kind of alluded to here, which is sort of the biggest hurdle for most organizations, getting that data together, and ultimately being able to analyze it comprehensively. So I think that naturally, if you think about all the things that go into that, you know, like you have to, you have to think about what data you need to store, nobody wants to store every single piece of data that flows through their organization, you need to think about what data you can store according to data privacy regulations, you’re not allowed to store all pieces of data legally, right. So there’s a lot that goes into that. And then ultimately, how you’re going to leverage that within your organization as well to actually make decisions. And I think that’s naturally leading a lot of the way like we work with a lot of mid market or enterprise level organizations, it’s naturally leading them into that data lake strategy, because there’s just simply no other better way to manage all those variables at the moment, or be able to really make sure that you’re, you know, maintaining the data set that you really need. And then yeah, internally, I think that you will start to see, yeah, it would depend on the company strategy. But yeah, you will start to see a I think used to also sort of a more of like an office of the CFO type level within organizations as as, as those types of software’s become more mature, but I think you’ll definitely see it a lot sooner. And with a lot more practical use cases, within SAS applications, you know, sort of in the near term.
Kristin Robertson 21:30
Great point, I also think that there’s really a two prong strategy to this that I see in the marketplace, there’s an appetite, need and desire to ensure that either banks or corporates have their own data within their data lake so they can self control what they’re doing from their own Gen AI initiatives in house. But I think when you look at the true value of where the market is going directionally, with generative AI is when you’ve got a multi tenant SaaS solution, whether it’s focused around Treasury payments around that area, the ability to have cross functional views of how you’re comparing to others in your marketplace, is incredibly powerful. But it again, it gets down to who has the most data to go into the data models and organizations agreeing to align on having, obviously not their sensitive data, but the general parameters as to how they’re comparing to others in the marketplace, assessable to the broader market. That’s where I think you’re gonna get an even more powerful analysis coming out of these machines.
Ron Chakravarti 22:27
I think maybe the only thing I can add, Craig, is this if I understood your question right? Now answer it this way, it’s increasingly going to be the case that that organizations, buyers of technology, whether it’s companies and their treasury technology, or banks is going to say that you can’t have something where the if you will, the if you think of it as three layers, the data layer, the process layer, and then visualization and analysis, like we can’t have something where everything is trapped within each each piece of technology. And that technology, the technology that if I’m going to buy from you has got to be open and interoperable. In that, at a minimum, the data has to be something that can go into the company’s data can ingest and exhaust data into the company’s data lake. So I might use my TMS because it’s fit for purpose for a certain set of things. But as long as I’m then getting the data out, I can build all the analytics I want, or deploy all the other tools that I buy and put in place, whether it’s for forecasting, risk management, you know, sort of getting FBN a to do a better job of the forward budget forecasts, etc. That’s going to be that interoperability and openness is going to be increasing the demand otherwise, you know, and I think that’s the right way to go that otherwise, I’m not going to buy that technology. It does, it speaks very well. But I’m in a closed loop otherwise.
Craig Jeffery 23:49
Excellent. I’ll ask you a question. That’s not to be answered here. But for everybody to ponder. If, if we have to get our data cleanse, to can’t we use AI and machine learning to help identify how to cleanse data? Yeah, think about that. Okay, so your head’s in a knot. Now, what about the human? What about the human element here? There’s there certainly changes the need to put this technology in place. We see AI and machine learning having a big impact in areas like auditing certain areas of finance. That’s, you know, taking out lots of jobs and creating other jobs and roles. What, what does it what does a treasurer need to do in terms of staffing? What do Treasury professionals need to do in terms of either gaining knowledge in this area? How do we how do we adapt to this rapid change in the use of AI and machine learning?
Ron Chakravarti 24:41
For all organizations? It’s, I think you’ve touched on a key point, it’s about how you think about your talent. And again, thinking about companies for the moment our clients, the I’d say that it’s not surprising this correlation happens that the more advanced companies that are more advanced Some Treasury tend to have the capacity to think a little bit more about the future. And what I see happening is this very deliberate, rather than accidental. As people leave a joint strategy towards, we need to have Treasury be data savvy, we need to have the Treasury team be technology savvy. And we’re going to have a deliberate strategy of how we bring in to this isn’t about having the traditional Treasury tech resource that’s in the firm that supports our TMS and other implementation, it’s within the within the department itself, we need to have people who attack and particularly data savvy, otherwise, we’ll never be able to figure out, you know, you’ll have the AI experts, because they’re always once in the firm, all big companies are using AI and the business even if they’re not in finance, and Treasury, and then you’ll have the subject matter experts. And the Never the twain shall meet, because they’re talking past each other that talent management and building that is, is it’s what’s going to be really important. I do see that happening. So that happening at banks as well. I’m just talking about companies for that. Is that
Craig Jeffery 26:01
like a new business analyst role to bridge the gap between the business and the AI experts? Or is it something else? I
Ron Chakravarti 26:09
think it’s I think there’s an element of that. But it is about having people who come up the Treasury ladder, who are not finance experts in the in sort of in the traditional there are this might not be CPAs. But they are they’re good accounting or sort of finance. But from the start, they’ve got a training and a background and data and technology, right. And if you look at what people who are graduating now, not just in finance, but other disciplines, it’s increasingly a part of kind of the education they go through. So it’s part of bridging, but it’s also partly having people who really are not just the analysts, but actually think think about the possibilities. As they start thinking about their initiative. As they start planning and doing their initiatives,
Kristin Robertson 26:56
I think it’ll end up taking away a lot of the manual processes that are being used today, where they’re not necessarily critical thinking, where you’re spending time and resources, putting fingers on keys, I think that that is a thing of the past over time with system to systems integration with using you know, generative AI and machine learning to help cobble all the various data points together, getting all of that time taken away and focusing on really the value added humanity side of things, making sure that the data is in the parameters are not biased, making sure that you’re you’re vetting the various scenarios. And it gives you maybe some ideation that you wouldn’t normally think up on your own. But you still need to weigh the pros and cons of everything that’s being produced, that you’re getting out of the systems and have the human intelligence and interpretation factor, whether it’s data scientists, the business acumen, the two talking to each other together to get the most value. That’s where folks are going to be staffing up in the future, not necessarily for the more manual tasks for my view.
Ron Chakravarti 27:57
I mean, if I can add to completely agree and as my colleague David Tao says that people are very bad at repetitive, boring tasks, right? And the power of this is going to be in a completely agree in automating the repetitive tasks, whether it’s about the data cleansing or the world based on your cash position and your forecast, what’s the most likely trades that you need to do even though the human then does it, but then freeing the human, you know, people to make that to say, Okay, what really should I do? So that completely agree with you?
Jon Paquette 28:29
Yeah, I’d say that this concern comes up all the time, will AI take away jobs? And I think the answer is yes, in certain industries, it certainly will take away jobs. I don’t think there’s any way to get around that. But I don’t think Treasury necessarily needs to worry about that. I think Treasury has probably naturally insulated from it a little bit in the fact that treasuries have never had big teams to begin with, you know what I mean? Like, it’s always two to three people working within a Treasury Department. So it’s not as if there’s this massive department where all these efficiencies are going to suddenly result in big reductions in the need for staff. And if you look at technology, historically, adoption of technology and treasury, you know, the the TMS when the TMS first sort of came out, it was a tool for automation reporting within treasury, Treasury has got more efficient, right. And a lot of Treasury Departments hired additional staff, because the use of technology allowed the entire role to change, it was no longer the Cache Manager, now you’re managing risks and capital structure variables and all these other aspects. So it actually made the, you know, made the role of treasury I guess, more prominent within the organization. And if you look at the kind of the requirements on Treasury these days, the expectations have totally changed in terms of what’s expected out of a cash forecast, the role that you play within the business playing a more strategic role. You know, even we talked about data if you if you think about Treasury’s role in data, they’re probably collecting more data elements in any other department within the Office of the CFO. They’re looking at AR AP, bank data, GL data, payroll data. So I think that I think of AI is more of like the tool that is going to arm the treasurer’s to meet those expectations. I don’t think it’s going to take anything away. I think it’s going to allow treasurer’s to be able to kind of take that next step up and become even more strategic within their within their companies.
Craig Jeffery 29:59
I’d like to use your words there. Yeah, so let’s shift over to development cycle and expectation. So we’ll start with, we’ll start with Kristen. So, you know, each generation of technology allows for faster and faster development. If you look at the new technology, we’ve used the term cloud native. So think of Amazon Web Services, Microsoft Azure, the development cycle, and people are building in those type of platforms, the development speed tends to be five to eight times faster. That’s a, it’s not 10 times faster, but it’s significant. And that’s making a huge difference. So development cycle times tend to shift. What are the implications for this? For organizations, we’ll go Kristin, and John, then we’ll move to a speed round and I’ll start with Ron. So Kristin.
Kristin Robertson 30:47
I think the the implications of going to platform as a service with containerization. And being able to actually make rapid changes that only impact a very small section of a much broader system is critically important to the overall industry’s ability to flex and flex rapidly. It also ensures that you’re doing full quality assurance controls, because when you look at historically, whether it was on prem or more software as a service, in order to change one little area, you would have to regression test the entire code base, we’ve gotten away from that technology with platform as a service to allow, again, focus on rapid change in the area that only matters most for that particular topic, allowing me to faster, better sprints, more quality code, and basically transfer my transformation across the treasury workstation space.
Craig Jeffery 31:40
I think everyone listening should replay what Kristin said a few times the containerization aspect and making things being able to change them on the fly without having to bring the whole system down. It’s, it’s kind of like the difference between, you know, a mechanic and a surgeon, you know, the mechanic and the surgeon say, well, we replaced parts in a car and the surgeon is like, I do it when it’s, it’s running down the highway at 60 miles an hour. And so that’s that’s a very, very significant difference. And I think that needs to invade our, our minds as to how we think about so that was an excellent description. So replay that. Jon, any thoughts?
Jon Paquette 32:15
Yeah, I mean, I agree, I agree with everything that Chris has said as well. But like, it really has changed the game, I think, for software development, and you’ve seen such a rapid ability to develop platforms so much more rapidly, you know, just SaaS in general, being able to multi tenant SaaS, being able to build something once deploy universally across all your clients, and through Infrastructure as a Service be able to do that through like rolling deployments that don’t even result in platform downtime, while you’re upgrading your platform. It’s just a huge difference from the classic model of how sort of software was maintained overall. And I think like also just being able to for SaaS companies being able to launch new products, and then just automatically scale resources through cloud platforms without having to worry about that infrastructure. That obviously takes a lot of the complexity out of a lot of those actions as well. So I mean, we’re even to the point where we can do weekly releases, if we want it was if we can, we can significantly increase our release cycles without it really putting any additional burden on the on the business whatsoever.
Craig Jeffery 33:06
So how does this impact how organizations look at the technology? Right, we’re familiar with SaaS, you know, we had older terms, you know, installed than we had hosted. And then we had Software as a Service. So there’s now software as a service platform as a service or cloud native. How should that this change had technology impact how corporations look at their technology providers?
Kristin Robertson 33:32
I’ll take that to start, I think it’s important to understand from a technology provider perspective, the DNA of the organization that you’re working with, and where they’re at in their own technological debt evolution as well. So if you look at organizations in the marketplace that have, you know, some systems that are 20 plus years old, really questioning, you know, what their cloud journey is how they’re taking those same features and functions forward, versus a market disrupter that may have gone to market day one, completely cloud native platform as a service, where they’ve built from the ground up to suit that infrastructure. A lot of times those market disruptors only have a very small fraction of what’s needed from a functionality perspective, but they’ve got the right technology approach day one. So it’s really diving deeply into a the organizations that have been in the market for a long time, where are they at in their strategy in their evolution? And be if you’re looking at a market disrupter, can they actually accommodate what you need and and to solve for your your pain points?
Jon Paquette 34:32
I agree with that completely. But yeah, I think that it’s, it’s a it’s a good way for organizations to address tech debt, I guess, get some efficiencies in place where like, you can’t lift and shift sort of this legacy tech debt overnight, obviously, like we’ll work with companies that have, you know, 20 plus on prem ERP systems, if you think about that, those are localized systems, siloed processes, siloed data, no way to sort of manage that organizationally across the entire business. But if you throw a SaaS product on top of that and integrate with all those on prem system As if it has the capabilities to do that, suddenly you’ve unified that whole environment from a from a control standpoint, from a data standpoint, from a process standpoint, workflows. So it’s a good way to, I guess, plug those kinds of gaps. In the meantime, I think it’s hard to argue that the on prem model will win out long term, obviously, it seems like everything’s moving cloud cloud ERP migrations and things like that. But that obviously takes some time for companies to to address.
Craig Jeffery 35:23
So the question for everyone to think about is what model and what decisions will enable you to make investments in technology that are an asset that increases in value, versus one that static or decreases in value over time, think thinking about that will have a big impact on what you do. This brings us to our first and only speed round of the day, just gonna go over the rules we’ll start with, we’re gonna start with Ron go down, and then we’ll, I’ll pick a different person to start each time, we just have a few questions. So you must make a choice. And prioritize one over the other as uncomfortable as that will be lead with your choice, and then give a reason for the other, you’ve got about 20 seconds to go. Now and in five years API first are all of the above. So you know, every connection method is all the above API first, or all the above?
Ron Chakravarti 36:13
All of the above now, of course, but in five years API first caveat is that there are going to be a lot of companies that are large, that have legacy structures. If it ain’t broke, don’t fix it. And it’ll take them longer. But increasingly, from five years on, it’s going to be API first.
Kristin Robertson 36:31
I agree. I do think that API first and five years probably going from you know, 80/20 rule route flipping that around. So five years from now, hopefully, the connections would be 80% of the market API first. But you’re gonna have laggards and old systems out there. And still, it’s going to be a long tail to getting 100% on board for API first, in my view.
Craig Jeffery 36:52
Okay, any disagreement? Jon?
Jon Paquette 36:54
I’d say all of the above even five years out, there’s there’s a lot of tech debt out there. There. There just is I mean, if not to pick on the banks here a little bit. But the way the banks infrastructure sort of built in, you know, to API first, you really need software that’s built up with microservices and interacts with each other through API’s, then you open those API’s more broadly to the outside world, right? And to think about everything shifting in that direction in the course of five years. That seems like a little bit of a tight timeframe for me, to be honest.
Craig Jeffery 37:21
I’m glad I didn’t say API only your or all the above. Alright, so Kristin will start with you with the second question. Thanks for Thanks for your responses, being quick with the answers and laying it out. So which will be the most transformative and how software as a service, or platform as a service, like Amazon Web Services, or Azure go,
Kristin Robertson 37:44
I would say Platform as a Service will be the most transformative and that via microservices, being able to embed the experience will really drive market adoption from that seamless for frictionless approach going forward.
Jon Paquette 37:57
I’ll go, I think I’ll go, I’ll stick with my guns. This is a tough one too. But I’ll stick with my guns on this one, I go Software as a Service, just because I’ve seen it transformed. So many businesses already, like being able to get into the cloud, like I said before, break down processes that are siloed. And there’s no way out of those silos, besides having a cloud based SaaS application has absolutely transformed companies. And the progression to that, you know, is sort of like getting on universal, you know, ERP systems in the cloud, being able to for like a Global Fortune 100 company, being able to to the for the literally the first time, the existence of the business, unify the operations across the entire world to a single platform single processes, single functions. That’s, that’s really transformative. So I’d probably go with SaaS on that one. Okay,
Craig Jeffery 38:35
So Ron, who’s right?
Ron Chakravarti 38:37
I have to go with PaaS, because SaaS for all of its advantages. In the end, you’re going to run out of capacity, you’re going to run out of runway for what you’re trying to do. And if you think about platform as a service as the end stage, you might start but sighs but that’s where you got to get to otherwise you just going to be again, in a different silo?
Craig Jeffery 38:57
All right, excellent. So final question. And which matters more about data speed, or the amount of information so which matters more about data, how quickly you can get it and access it, or the richness of it.
Jon Paquette 39:10
I’m gonna say, from the purely the treasury perspective, its speed of data, because I don’t think you need that much information in the data to make good cash management, treasury decisions, you just need data to be flowing frictionlessly frictionlessly into your business, to be able to get it as as much as you need it to make those decisions. So I think speed is really the key there for treasury.
Ron Chakravarti 39:29
I’d say it’s data. I mean, I know you said data is the speed of or amount, I’d say it’s data, getting the data getting as in having completeness. You gotta you gotta have the data. Most GM, most organizations don’t have a data strategy in place. They’re not thinking about the data, how quickly you get it into the point, I just found it. If you’re not going to be APR first and the speed of the data data doesn’t really matter. If you have if you don’t have the richness of data doesn’t matter how quickly you get the data, you’re not actually going to be able to do anything even from it from a treasury perspective. To have to be able to make sure that he was spotted that that you’re doing the intercompany posting right, that particularly as you if you’ve been if you’re a big company or operating in regulated markets and free markets, you’ve got that particular little deterministic sort of code that tells you why you were able to sweep in and out of China not so. So it’s a fun data.
Craig Jeffery 40:21
Okay, so is the amount of information I thought you’re violating our rules, but you hadn’t. So error on my part, but excellent answer. So now we have another tiebreaker. Kristin, this time, you get to say who’s right?
Kristin Robertson 40:31
I’m gonna have to agree it’s about the data. Because if you don’t have the right information, it doesn’t matter how fast it is, you’re making decisions without a full view of what’s going on.
Craig Jeffery 40:41
So if I get the Wall Street Journal delivered to the bottom of my driveway, three days late. That’s just another story. Yeah.
Jon Paquette 40:51
Every article in the Wall Street Journal was given the headlines to get the points.
Craig Jeffery 40:55
This, this is awesome. We had we had a prep session. And I tell you, we had to control ourselves to stop from going on. This has been a lot of fun. All right, so now we move to data and processes. And who who starts first on this one? How about Kristin? Why don’t you start first on this one as well, the total amount of data is doubling every two years. This is a massive explosion of data, we’d have lived in a time where we couldn’t get the data. And now we have a tremendous amount. So what’s the potential and power of data and insights and the power of analytics? And keep this quick to we’re going to give everybody a quick shot? And then we’ll do the future look?
Kristin Robertson 41:31
Well, I think the power of data and insights is going to be critically important. But I think your question of who’s cleaning and making sure that the data as it’s growing is, in fact, accurate data is going to be critical going forward. And having machines police the machines is kind of an interesting concept that you need humans to interpret and data scientists over time. But the insights are going to be invaluable as the data grows over time, as long as it’s accurate data.
Craig Jeffery 41:56
All right, Ron, would you like to?
Ron Chakravarti 41:58
It’s getting it, getting it in and getting it clean and getting it good? Is this what’s critical?
Jon Paquette 42:04
Yeah, I mean, I agree to it’s I mean, you can’t a lot of these technologies we’ve been talking about today are no good without the data, obviously. And you know, being able to get that in and have it be clean and making sure it’s the right data and organizations thinking about overall that data management strategy for how they can collect what they actually need, and doing it in compliance with local regulations and things like that is yeah, it’s critical precursor.
Craig Jeffery 42:24
All right, so we’re gonna have to shift to our last question. Due to time constraints, I know we could go quite a bit longer, but we are going to be compliant with the AFP. So a future look, I’ll start with John and come back towards me with a so I’d like to hear what you what we should expect to see over five years and 10 years with regards to treasury technology, and feel free if you’re the second or third one to acknowledge what someone else said. But let’s not repeat exactly the same thing. You can acknowledge it. But let’s go Jon, Kristin, and then Ron.
Jon Paquette 42:56
So I think it’s all framed up around this, this kind of point that I made the beginning of the changing expectations around software, where people really aren’t looking for automation and reporting anymore. They’re looking for insights, and ultimately triggering that into actions. And like I said, it’s why a lot of the technologies that are out there today are getting the attention that they are. So I think you’ll see providers focusing on developing their products in that direction, launching products that are more geared towards helping customers get to those decision points. And they’re going to do that through things like incorporating in AI, which can help a company make it much better, quicker, more validated decision a lot earlier, and then translate that into an action, they’re going to do that through things like embedded finance, because ultimately, when you get to that decision point, you then want to say, Oh, now we have to leave the solution and go launch some money market fund portal and Keenen investment. And then we have to go to the banking portal, and kick out a wire to settle up that investment. Now you want that to all be sort of integrated into the into the application itself. So that whole process is completely frictionless. And it’s going to have to do a lot with speed of data. So I mean, so API’s you know, that’s that’s sort of the way that people are thinking about that right now, it could be that new newer technologies emerge that allowed data to come in a lot quicker. There’s a lot of like banking as a service products out there in the market right now that are pretty interesting. That could maybe help the banks address, address some of that tech debt and innovate a little bit faster. So but I think you’ll see people pushing these same topics, but under that more unified lens of like making the treasurer better at their job, essentially,
Kristin Robertson 44:21
I think you’re going to continue to see the smaller firms that are out there companies in the US leaning more to their banks to help them from a treasury workstation perspective, as banks are getting more and more sophisticated with what they’re providing. And as organizations are rationalizing the number of banks that they have, and that’s frankly, where they’re going to focus to get stickiness in the commercial space. And that you’ll see that Treasury workstations will continue to embrace Platform as a Service will continue to focus on the larger corporates who have very complex global multi bank environments where they need something that’s really specific to their organization, but being able to ensure that they can replicate their data they can use that data within their four walls for their own generative AI and purposes to keep their culture their perspective. Their optics will be really where we’re going from from a futuristic perspective.
Craig Jeffery 45:11
Thanks, Kristin. Ron, you started us off and now you can finish it.
Ron Chakravarti 45:14
So Kristin use the term frictionless first. And I agree with everything that she said. And that term friction strips frictionless that John also use. And I do agree with what she’s laying out that you’re going to have this larger companies are going to want to be able to transact with that’s the term from their tech stack with the financial services industry, right? They’ve got a structure, they’ve got architecture, they’ll transact from there. For midsize and smaller companies. The investment budget is going to be hard to justify even if it’s SaaS, etc. And I think they are going to go towards expecting and getting from banks who capable banks, a lot more of the services sort of it’s almost like the TMS industry and a number of other sort of tech industries that have that are in the treasury space are going to reform along. completely embedded humans are making this sort of like the smart analysis with the systems taking care of it. This is about 10 years ahead. Versus midsize and smaller companies, much more willingness and trust with the right financial institution to rely more on the tech stack they’re giving you.
Craig Jeffery 46:20
Excellent. So thank you Citi, Finastra, and TIS. Really appreciate it. Let’s give them a round of applause.
Ron Chakravarti 46:26
Thank you for coming.
You’ve reached the end of another episode of the Treasury Update Podcast. Be sure to follow Strategic Treasurer on LinkedIn. Just search for Strategic Treasurer. This podcast is provided for informational purposes only and statements made by Strategic Treasurer LLC on this podcast are not intended as legal, business, consulting, or tax advice. For more information, visit and bookmark Strategic Treasurer.com.
Access your definitive guides to treasury technology. Researching new treasury and finance technology can be overwhelming. Strategic Treasurer has stepped in to help. Explore our definitive guide to the treasury technology landscape and discover detailed, data-based coverage of:
Treasury & Risk Management Systems
Supply Chain Finance & Cash Conversion Cycle
Liquidity Enterprise Management