The platform mindset, a fundamental shift in thinking, has gained prominence in the fintech community and disrupted the way banks do business. Traditionally, IT systems at banks were built in-house, and were custom and comprehensive. This gave banks complete oversight in the security and design of their systems, but it also made their IT difficult to change or improve.
A platform mindset involves building IT in modular units that can easily be swapped out or outsourced. These units are not necessarily developed in house. In such a model, the platform provides a capability that is supplemented by technologies from an ecosystem of partners.
To learn more about the importance of the platform mindset in IT management for fintechs, Payments Journal sat down with Tim Sloane, Director of Payments Innovation at Mercator Advisory Group, and consulted a recent e-book by Lumen, Embrace the Change: Charting a Course to the Future of Fintech.
Legacy IT Approaches Vs. Cloud Architecture and APIs
Until very recently, established financial companies built proprietary, monolithic IT systems for banking, payments processing, and capital management. These applications were often hand coded. Their IT architectures relied entirely on in-house, dedicated resources. Companies built up their IT infrastructure as they became more successful by adding data storage centers and physical networks.
Application program interfaces (APIs) were less common. APIs are a type of software interface, offering a service to other pieces of software.
Over time, certain financial companies tried out new computing architectures and even created internal APIs. As Tim Sloane recounted, “Back in the ’70s, and ’80s, APIs were used internally, and companies developed their own custom APIs. For example, Fidelity created their own APIs to be able to link all their different systems together.”
Recently, API use has proliferated. Sloane noted that this came to a head when, four or five years ago, Visa and Mastercard started to publish APIs that would enable third parties to access their payments networks. That made way for fintechs to develop innovative applications which could now access the payment infrastructure.
Over time, banks may have embraced the cloud for their internal needs, but only superficially. The Lumen e-book noted that, “Hand-coded applications would require rearchitecting to make use of containers, micro-services architecture, and other aspects of the cloud.” A fundamental shift would require disruptions that most banking customers would not tolerate.
In contrast to legacy financial companies, many newer companies use a cloud-based architecture. APIs can integrate new capabilities relatively quickly, as well as grow IT more organically. Sloane summarized this nicely: “If all of a sudden a business triples its number of customers, traditionally it would have to first increase the amount of hardware it has on premise. With the cloud, a company can simply buy more remote disk space as needed.” Conversely, if, for example, a holiday surge ends, a company can reduce its amount of CPU and disk space. The result is that costs are more directly aligned with utilization.
A New Model: Microservices Architecture
According to Sloane, the old style of IT architecture was to write one set of code that does all of the things a company needs. In a microservices architecture, however, programmers write a small bunch of code for each function to execute and orchestrate those functions with APIs. Each bit of code provides a “microservice” and can be updated and swapped out separately. For example, a company with a microservice architecture might outsource authorization of resources to a third party.
Microservices enable significantly more flexibility, especially when implemented in the cloud. However, as Sloane notes, financial services have traditionally done monolithic code for a reason. “The developers writing financial services products are monitored very closely to make sure that you don’t have a coder putting a back door into the software. And they have tight quality control, so that when the software is released, the company can be sure that it’s going to work properly.” Quality control, as well as the oversight by regulators, has driven banks into that monolithic IT architecture.
As a result, it’s been difficult for financial services companies to adopt a microservices architecture from a software perspective, but also from a regulatory and process perspective. Changing to a microservices model requires retraining developers, redesigning processes, and redoing systems.
A platform model with microservices and cloud computing has risks. “Cryptocurrency platforms out in the market today were built fast with microservices architectures. And the hackers and criminals are having a field day with crypto markets,” Sloane notes. Nevertheless, microservices could lead to more flexible and dynamic options for financial services companies in a variety of ways.
The Platform Mindset
Part of the issue is mindset—specifically, moving toward a platform mindset. As the Lumen e-book notes, “By constructing a platform – assembling and integrating various technologies from the right providers – fintech players can build a bridge to the future without disrupting the present.”
Lumen notes that the fintech and media streaming businesses have a certain similarity in this regard. “Both are highly distributed in nature. Both are data intensive with multiple applications handling different aspects of the customer interaction. They assembled a platform they control, using a set of partners who provide the best-of-class components.” Netflix, for example, does not produce all of its own content, nor does it have all of its IT equipment in-house. Instead, they have a platform, which makes use of external assets and technologies, but unites them in a way that customers love.
According to Lumen, a good platform includes data operations that are cheaper and more flexible, and excellent customer service. For example, Lumen has implemented a better bill pay experience for consumers by creating a platform that integrates a bunch of different technologies from other companies. On the platform, customers can select from any number of banks or a number of cards to pay bills and get instant feedback that the bill has been paid. This makes use of cloud computing, microservices, and APIs.
A Hybrid Solution
For many companies, the best strategy is to shift to a hybrid environment that combines internal resources with one or more cloud providers and other outside resources.
One example of a hybrid solution could be taking a monolithic code and renting cloud storage and computing power to run it. If that company needs more horsepower, they can add it quickly. They can also distribute those systems geographically, so there is no time lag for international businesses.
Another type of hybrid solution involves using cloud computers but storing the data locally as well. This is an important issue in national security. For example, in certain countries resident data legally has to remain in the country. A hybrid IT solution makes it possible to comply with this but retains a certain amount of flexibility with cloud computing.