‘As far as next gen cores are concerned, fintechs lead the way’: Fiserv’s Frank Sanchez

The following was produced by Tearsheet Studios. We worked with payments provider Fiserv to create a podcast series about open finance and the work of empowering fintechs, brands, and FIs to collaborate and innovate together.

In our second conversation in the series, we’re speaking with Frank Sanchez, vice chairman of Fiserv, and co-founder of Finxact. We talk about the promise and challenges with next generation cores, how fintechs are leveraging them in pushing open finance forward, what sets them apart from legacy banking cores, and what innovations we can anticipate in the near future.

Subscribe: Apple Podcasts I SoundCloud I Spotify I Google Podcasts
The following excerpts were edited for clarity.

Frank Sanchez, Fiserv: I'm Frank Sanchez. I am currently the vice chairman of Fiserv, and managing director of the Finxact subsidiary, which I started in 2017 as a founder and CEO. Prior to that, I've been in the core banking business for 44 years.

With your history in mind, let's zoom out and talk about how you got to where you are today – I know this is not your first rodeo.

That's exactly right, it's not my first rodeo. In 1979, I started a company called Sanchez Computer Associates along with my brother and dad – we designed and brought to market the first online banking system. It was successfully installed in over 200 institutions globally. We went public around 2000, and sold the company to Fidelity Information Services in 2003. I went along with the sale and moved from Philadelphia down to Jacksonville, Florida, where I reside today. 

I was responsible at FIS for a division called Enterprise Banking, which was in the top 150 global banks; I was also responsible for all product development across the 14 cores that we had at FIS. I retired in 2011, and joined a fintech startup in Silicon Valley called Zenbanx along with a good friend of mine, Arkadi Kuhlmann, the former chairman of ING Direct; we sold that company to SoFi in 2016. At that point, I was sort of asked by executives in the industry to take one more lap around the track, if you will, and build a next generation core. So in 2017, starting from scratch – tabula rasa – we bought new computers, new desktops, and went off to build the next generation core. Now, five years later, it was acquired by Fiserv, and is successfully installed in about 20 institutions. We have eight or nine other projects underway, so we're off and running.

We've been talking to Fiserv about the evolution and possibilities of open finance. Let’s talk about the role next generation cores like Finxact facilitate in open finance.

The legacy cores were really responsible for one function: the basic accounting of transactions and record keeping for customer accounts, reporting to the general ledger, and producing customer statements Attempting to integrate those cores into an ecosystem where you might have additional content, other applications, integration upstream and downstream, was very difficult to do. This next generation of cores is designed to be part of an ecosystem – not to stand alone and perform a single function.

What factors are driving the move towards open finance and open ecosystems, and along with that, innovations like next generation core?

I think people were looking for more of an end-to-end integration where you might have a frontend application for a dentist, or some other vertical application that's responsible for providing functionality to a dental practice. They were looking for practice management systems that integrated the financial function right into the system, so that the payables immediately produce payments into the payment system, and the receivables immediately go out and debit accounts –so much more of an end to end integration.

There is also more flexibility; and in order for it to really work, it requires a move to more of an API driven architecture where products like ours publish the interface, then can be readily consumed from other applications without having to go through a significant amount of development from either party. 

This is very similar to what people have been doing in integrating with other SaaS products like Salesforce where it becomes part of a broader ecosystem. On the functionality side, we're seeing more of a hybridization of products where they're not only dealing with cash, but also Crypto, or securities of another nature. And this means that managing a customer really means that you need to manage them over a broader spectrum of products.

What does the next gen core do that earlier ones didn't? What problems does it solve?

Firstly, the next generation cores are cloud based. Because you are opening them up for other applications to utilize, you can't really understand what the scale may be. And, there can be busy days and slow days; you're in an e-commerce type of an environment, and you're capturing much more about a transaction than a legacy core. 

Legacy cores are designed to be accounting engines, whereas the next generation cores are tracking much more information around transactions that can then be used for customer analytics and fraud in a very active and real time payment system. A next-gen core is enabled by what we call RESTful APIs, which means that it's as accessible as any web page would be through an HTTP request. It's a much simpler environment to integrate with. 

However, that brings a lot of additional burden and demand for security, because now you've actually opened up the accounting and record keeping infrastructure to a broader community. When I got involved in Finxact, it was important that we focused on just building the best real time position keeping platform. We took the vertical stack where the legacy applications were – the frontend, the orchestration layer, the accounting engines or system of record, the reporting systems – then split that out and said, ‘We're going to do this one piece of this thing very, very effectively, which is a system of record. And we'll allow other best in class applications to deal with the user interface, channel applications and risk, fraud analysis, and customer analytics. What we're doing is broadening the universe of capabilities and allowing financial institutions of any kind, and even non-financial institutions, to plug and play components – but when they  hit the ‘Post It’ button, they have a very reliable, real time position keeping platform behind it that does that one thing very, very well.

How would you categorize this last lap, as you described it, compared to some of your previous experience at building banking software? Is building banking software today a much bigger undertaking?

That's a good point. It is and it isn't. Building banking software today is a bigger undertaking in that the level of sophistication of the client and financial products produced is always going to increase over time; everybody starts by saying, “We just want a simple X”, but it never actually turns into that as you get into the product, because financial institutions want and need to differentiate themselves. So you have additional complexity there. 

There are clearly more payment options today. If you're connecting to payment systems, you’re facing a global audience, so you have to consider things like multi currency and multi language; though the language isn't as much a part of my application, it is part of the stack, and there's much tighter control over authentication and encryption, with all the security aspects and fraud.

Clearly the list requirements has increased over time. It's offset, to an extent, by the hardware being more abstract; you're dealing in a cloud – a container – so you're not coding for a particular piece of hardware. If you want to get another piece of hardware, you don’t have to actually change your code.

The development environments themselves are much easier with the modern languages. We use a language called Golang but all languages in this tier are much easier to code in and much easier to debug – the development process itself is facilitated. The tools that allow us to develop and test are facilitated. A very big part of the development process on any product is being able to clearly articulate specifications, to be able to model your specifications, and then to be able to set up test cases that can hopefully run as much as possible automatically, so you can do regression testing every day. 

The actual toolkit to allow technology development has improved substantially, even over the last 10 years. Especially getting on a cloud, where everybody can join up in an instance and have a copy – us, our customers, the developers – and we can instantly create those containers on a cloud, and again facilitate the development process. 

There are pluses and minuses. As we get into this, I think net net it's easier to build applications today, but somewhat offset by the fact that there's much more scope to build towards than there was. 

Can you talk about some of the attributes that define a next generation core? 

The table stakes are that it is a cloud based application. Ours, and others, were designed from scratch, so every line of code was written from scratch, and written for the cloud. There are facilities that are available by the cloud providers that we want to take advantage of around security, database, elastic scalability; writing for the cloud allows you to do that. 

That’s different from running on the cloud – anything can run on the cloud: Microsoft Word runs on the cloud. The code was written for a PC, but it runs in the cloud, and legacy COBOL applications can be lifted and shifted to the cloud as well. We need to differentiate running on the cloud versus designed and built for the cloud, because one is a hosting facility, and the other is taking advantage of the infrastructure that these providers offer us.

I believe RESTful APIs are another part of the table stakes. The way an application or any cloud based system of record operates is through that HTTP protocol, because that's what allows us to be accessible to the outside world without writing custom point-to-point interfaces.

Being real time is critical. The world is moving to 24/7 availability; the idea of a bank being limited, or its hours being limited, by a branch network is archaic even now, so you have to eliminate the notion of time zones being a factor or expecting your customers to be in particular time zones. The updates to the accounts, or the positions in the case of fintech, have to be real time, yet still have to interface with and accommodate the batch world because a lot of the payments are coming in through ACH or X9, which are both batch protocols. The accounting updates to the corporate general ledgers are on a daily basis, which are batch updates, so you have to accommodate both the real time nature of the channels that you're operating with, and the batch nature of a lot of the systems that we still have to interface with. Real time is an important consideration. 

Another important consideration is not having any kind of a day end batch, or day end process that in any way stops the system from being 100% available to its channels. That's something that we spent a significant amount of time designing into Finxact: an event based architecture. Everything that happens in Finxact that's based on a time series like an account maturity or rollover, interest posting, statements, is all driven off of event calendars. Every position, account, party, or customer in the Finxact system has an event calendar showing that  the next thing that happens will happen at this point in time. While we were there, we even eliminated the notion of a day being the smallest unit of measure in banking, so now we can offer 27 minute loans if we want, or one and a half day term deposits, so everything's kind of moved to a time based scenario. That's facilitated by the eventing and the calendar side of it. Those are all important. 

Zooming out a bit, what are some of the ways financial institutions and fintechs are leveraging next generation cores like Finxact to move financial services forward?

The first thing we're seeing is that they’re developing new products with things like multiple positions. If you look, for instance, at the account offered by One, a fintech acquired by Walmart and one of our larger install bases; when you open an account there, you immediately get a savings position, a transaction position, and a line of credit, and they're all under the same account. Funds are moving depending upon the payment source and customer limits. The funds are moving in and out of those positions seamlessly to the customer. We also see other banks and FIs providing similar capability. It's much easier for customers to manage your cash flow when it's seamlessly sweeping back and forth, and it’s all under the same account and rolls up to the same statement.

The connections to the payment systems are also real time. The settlement period of holds, term holds, and the process of settlement will continue to move towards more real time. It's not real time today on many of the clearing interfaces, not by nature of the next gen products, but by nature of the payment system itself. We have a lot of initiatives that are underway with different payment systems to move that way, which is another thing that you’re going to see that's a differentiator. 

The last thing is integrating into an ecosystem. For example, someone may have an application that reads small business general ledgers and we can integrate directly into that, and can look through their payables, evaluate them immediately and finance loans for them; or look at the receivables and immediately issue ACH requests for that; so without some kind of additional steps in between, it becomes a seamless integration. 

It's becoming more efficient to be on a next gen platform than on a legacy platform. Because the elastic scalability that cloud providers offer us says that we don't have to have a crystal ball of what the system is going to look like, from a scale perspective, three or four or five years from now for our own procurement requirements, and we don't have to scale to the worst day and pay for it, because the elastic scalability lets us scale on demand as we need it. So even though, arguably, the cloud providers are making money, the efficiency that is gained by the elastic scalability is substantial. Supporting these systems using modern languages and modern development environments is substantially more efficient than supporting legacy systems with, you know, resources that are using technology that's hard to acquire, and hard to pull somebody out of school and learn. So, with the whole transition of the tech industry in the last five or six years in language, infrastructure, and platforms, it is just more efficient to use new things than to try to keep old things running.

What role has the flourishing of fintech and new products and services played as an input into what a next generation core looks like? Is there a connection between the renaissance we've seen in fintech products and services to the cores?

That's a very good point. As far as next gen cores are concerned, fintechs led the way. Banks were constrained in what they could offer. The larger banks have been surrounding cores for probably 15 years, where they're actually trying to de-content the cores, because they were intractable even though they're the basis for a lot of the features of the products that banks are able to offer; trying to change them and that whole system was very complicated. 

So, banks have developed these surround strategies, where they now have logic and data across multiple platforms, and have a big reconciliation issue anytime they make a change or update something, because it's residing across multiple platforms. As a result, they were unable to really be innovative. The legacy providers, frankly, weren't in any better shape. 

The fintechs came in clean slate and said, ‘I'm not encumbered in that way’. Frankly, there are some disadvantages there, too, in terms of knowledge of the regulations, but as far as flexibility and freedom to do whatever they want, they were able to do that. They thought out of the box, they created innovative products that customers were asking for. That, in my opinion, created the tension that forced not only the legacy providers, but the existing financial institutions to step up to the plate here and say, ‘If I don't compete my customers will leave.’ 

We clearly have moved past the day and age where a branch on the corner was the reason that somebody banked anywhere. It's all digital now. And so, in order to compete you have to have something competitive.

I was surprised when I got into the industry in 2017 and started this last initiative. I was actually delayed for two years, thinking ‘Banks never change.’ There were so many horror stories about cores, so you question yourself, ‘If I build it, will they come?’

For our last question, I’d love to turn our gaze outwards. When you look towards the future, what kind of innovations do you foresee? What kind of world do you imagine?

I think you're going to see a continued migration into real time settlement. There's no technical reason that this couldn't happen today with platforms like ours providing that, but it takes a while for an industry to transition over. 

I also think you'll see much more integration with artificial intelligence and decision processing to incorporate different financial services into peoples’ lives. For me, it all comes down to liquidity. From a customer perspective, they are looking for a true partner in their bank or FI. They are looking for systems that are able to take them down the road as they evolve their financial story and try to, if not accrue wealth, at least stay ahead of the curve. 

I think technology can definitely help you there. It gives you further integration and some of that is also driven by what I call ‘the fourth industrial revolution’, or The Internet of Things. Some companies have done a very good job analyzing things like emails, and taking and predicting what a customer or an individual may want or need. 

In my opinion, there's nothing that describes you better than your transactional activity. So in context, what you buy, where you buy it, and the trends there. That information is structured information and really paints a picture of an individual's financial objectives (or at least their direction and momentum). We’re going to see the utilization of that information along with the tight integration of liquidity into everyday lives. We are going to see a much tighter integration of all that.