Banks are finally moving to new core systems, leaving behind bulky legacy software running on Cobol and mainframes.
Bankers, who in the past have compared replacing a core with changing engines on a jet while it is flying, are now actively looking for new core systems, said Paul Taylor, CEO at Thought Machine, a cloud-native cloud banking system founded in 2014. The company doesn’t have to convince bankers that change is necessary, he said.
“We are in the confidence business, not the persuasion business.”
Thought Machine has 26 clients including money center banks such as Lloyds Banking Group, Standard Chartered, Atom bank, Monese, and SEB and as of a few months ago, JPMorgan Chase.
“We are the clear market leader in the cloud native segment,” Taylor added “We have never had a conversation that you should do this. Banks want to do it and want to know we offer a good platform to land on. A lot of banks are going to take it at a pace that is suitable for them.”
The software uses microservices and APIs.
“Everything is accessed through APIs, whether it is internal or external.”
That API-based design plays a key role in a bank’s transition to the Thought Machine platform.
“We want customers to have access to their own historical data and transactions. We ingest it into the ledger. Rather than just copy the data, we put it through the API. The API does all sorts of checks, so when the data arrives in the new ledger it is verified, clean and not filled with junk.”
Bankers have good reason to fear moving off legacy platforms which have been around for decades. Most core systems continue to work reliability but changes to support new programs or new methods of customer access, like mobile, can be tedious and time consuming to develop. Developers refer to the software in legacy systems as spaghetti code for a reason.
The complexity of legacy systems has been amplified by mergers and acquisitions. Taylor said one bank the firm worked with had 100 core systems, and it is not unusual for a large bank to have 10 cores acquired over years of M&A.
“Regulators at one point may have required Social Security numbers, and a later date decided not to. The data base accumulated complexity that is not really necessary.”
Taylor and several other key executives at Thought Machine came from Google and are accustomed to the cloud.
“The vast majority of software apps are developed in the cloud today; it is by far the most common way to develop software. Because we were early in using cloud, when the banks were ready to have the conversation, we had a product that was quite mature. The whole issue of cloud and security has moved on. Most banks are happy to commit to the cloud architecture.”
Taylor sees the pace of cloud migration accelerating in banking.
“It is happening faster and faster. We have seen multiple tier one banks moving to the cloud this year. I think that any bank that doesn’t have a cloud strategy or dedicated plan to move to the cloud is really going to get left behind.”
Banks are attracted to cloud because they can operate at a fraction of the cost of legacy systems, cloud offers greater resilience — a lot of legacy on-prem systems don’t provide much support for failover — and cloud offers speed of innovation and speed of delivery while legacy systems slow banks down, Taylor said.
On the supply side, several banking software companies provide a cloud solutions, although only Thought Machine and Mambu are cloud native, he said.
“You can’t move if no cloud solution exists.” Google is Thought Machine’s preferred cloud provider, but its software can run on any of the big three, and it can run on multi-cloud and multi-region cloud systems for greater resilience.