For the banking sector, years of digital upgrades have left many organisations with a tangled technology portfolio that can slow down operations. As many banks are businesses that deal with thousands of transactions, customer interactions, and payments in a single day – both internally and with their peers/ rivals, this complexity can make it difficult […]
For the banking sector, years of digital upgrades have left many organisations with a tangled technology portfolio that can slow down operations.
As many banks are businesses that deal with thousands of transactions, customer interactions, and payments in a single day – both internally and with their peers/ rivals, this complexity can make it difficult to be agile in a volatile market.
Many are turning to automation as a way of dealing with high volumes of processes. A study from Deloitte shows that 92% of advanced automation adopters use end-to-end automation as a part of their strategy now, or plan to within the next three years.
But legacy systems and complex, dated tech portfolios can be a barrier for banks looking to automate. In a recent poll by process orchestration firm Camunda, almost 80% of respondents from banking and insurance sectors said that legacy tech was keeping them from achieving hyper automation.
Take NatWest for instance, one of the UK’s “big four” clearing banks. It has more than 7.5 million personal customers and over 850,000 small business accounts in the UK alone.
Taking just a small, but vital, section of the bank’s operations – its fraud department – you begin to see how complex its operations are. NatWest employs 800 fraud contact centre agents, servicing multiple franchises across the organisation. Each agent uses an average of 14 applications per call, with over 60 applications used across all propositions. These include internal and external apps.
That is just a single department, but it is a complex picture of processes at just one bank. Across the entire organisation – and across the sector – process orchestration is becoming an increasingly vital component to modernisation efforts.
Camunda found 92% of respondents from insurance and 84% from banking said that the volume and diversity of components and endpoints across their company are increasing exponentially.
Organisations in banking average about 51 components/endpoints (the highest from all industries), representing an 18% increase over the past five years.
In this age of automation, how are banks adopting process orchestration, and what benefits is it bringing to the sector?
Repetition
Speaking at the recent Camunda Con event in Amsterdam, NatWest digital transformation director Milesh Chudasama outlines the challenges facing the sector.
“Especially in financial services, what we tend to do is build the same process and mechanisms for different channels,” he explains. “You will build a process for a telephony customer, then build a separate process for online banking, and another for mobile banking, to effectively get the same outcome.
“You are building multiple versions of the same process to perform the same action.”
He adds that one of NatWest’s aims is to bring in agentic AI across the organisation but says this is challenging because, due to the complexity already outlined, you risk giving the customer a different experience for the same journey, depending what system they use.
“We want to target the processes correctly and build it in a way where the system is orchestrated and standardised, so the (AI) agent is not having to think and the human isn’t having to drive the action. This means the customer can get a standardised output from us.”
Dutch multinational bank Rabobank found itself in a similar situation. Head of tech payments and savings Fred van Pouderoijen tells TechInformed that the Utrect-headquartered bank – which originated by serving farmers – had always been an innovator.
The bank was the first to introduce ATMs in the Netherlands and began its internet banking journey more than 25 years ago.
But there’s a lot of dynamic change in the sector, he adds, including more competition from fintechs, and increasing regulation.
“There is a lot of competition, and there may be consolidation to compete with global banks, but trust remains very important,” says van Pouderoijen.
To deal with this instability the executive saw that the bank had to change” and tech was “at the heart of this change.”
However, he adds that implementing change was challenging – not least , he admits, because the bank had built up “a tech stack full of monoliths”.
“Monoliths aren’t flexible, but we use a lot of microservices,” he adds. “There was no administration or automation to control these. To scale, we needed to find a way of orchestrating all of these different microservices.”
To automate processes, Rabobank partnered with German process orchestration provider Camunda. In payments, for instance, it worked with Camunda to standardise its sales processes, while reducing the number of components, reducing the complexity on the bank’s front end.
This, he explains, reduced the time to market by 30% which was “a great result” for Rabobank.
In its housing division, Camunda helped it to improve the customer experience by resolving incidents in Rabobank’s management system without the need for user interaction.
It also helped the bank identify risk categories, by adding events which allowed the firm to automate outreach, reducing the number of manual processes.
“Imagine you have 7,000 people working manually every day, and you manage to automate some of these processes,” he adds. “It saved each user 30 minutes per use case, which is time that can be dedicated to other services, while also removing much of the repetitive work.”
Collaboration
HSBC is a UK-based multinational bank with historic links to Asian markets – it’s one of the largest European banks in terms assets managed, and the third largest non-state owned bank in the world.
For an institution that has over $10.8 trillion in assets under custody, delivering straightforward services that are accessible anywhere and anytime is vital.
But transforming processes in an institution that has been in operation more than 160 years, and runs branches and subsidiaries in 62 countries, is challenging, asserts lead platform architect Dan Advidan during a talk at CamundaCon.
“I run a team that builds tool that makes tools,” he said.
“We are digital enablers, in the business of providing a service across the organisation.
This means building frameworks for mobile, security, content management, and process orchestration, so that the various value streams can focus on business verticals.”
This includes the brands that HSBC as group operates, which can include competing banks in the same market, and white labelled services, such as Marks and Spencer Banking.
“It is quite a challenging operating environment,” added Avidan. “As part of our digital transformation we moved from local solutions to a global solution that we can deploy regionally, that can be built fast and scale fast. But we also need to account for each bank as an individual unit, and that it meets all regulatory requirements.”
The digital transformation journey, with the adoption of process orchestration, has seen five environments deployed across 35 markets, for around 10 value streams – which total more than 1750 environments.
“Our architecture was a traditional API-led system, interacting with the backend mainframes and systems, but we found there was very little incentive to reuse solutions,” Advidan explained, saying it worked well, but he wanted to be more efficient.
In 2023, HSBC partnered with Camunda for process orchestration. The vendor helped Avidan and his team build a new journey orchestration and choreography platform split across two clusters.
It used Camunda 8 architecture to install connectors to its existing microservices platform, meaning the new architecture could be hosted in the same environment as HSBC’s existing IT stack.
Avidan’s team also began building a second cluster aimed at supporting high throughput use cases.
“It was important for us to have a single source of truth through the entire workload,” said Avidan. “The grand principle behind our approach with the new platform is what we design is what we implement, is what we monitor, is what we optimise.
“Business users collaborate with the development team, and share a workspace, where they can model the processes together. The very same process is what remains – it won’t go through any further transformation.”
The system also gives business users access to an interface where they can directly see how processes are performing and identify any failures, he adds, without relying on IT to mine logs. It also gives them access to additional analytics which can help them optimise processes to hit more APIs, said Avidan.
Starting small
The nature of banking, with a centralised customer base, means process automation can be significantly beneficial in terms of auditability and helping meet compliance needs, explains Camunda’s Kurt Petersen.
“Some banks are doing millions of processes every day and they need not only absolute surety of execution, but the ability audit that,” he adds. “Another area where Camunda can really benefit the sector is in terms of resilience and scale.”
Petersen, who has been with Camunda for little over a year, describes the firm – which started out as a consultancy in 2010 before developing its own process orchestration platform – as “a developer friendly, pro-code environment” meaning internal teams at large organisations can often work with it to develop the structure to meet their own process needs.
Most companies looking to improve their process orchestration start small, explains Petersen. “Most start with singular processes, identify opportunities to bring in value, then elaborate that out to full, end-to-end process over time.”
So, what are the top line benefits of orchestration, whether leveraging Camunda or one of its rival platforms such as ServiceNow?
Petersen sums it up by saying it is vital for banks to “have the capability within a process, flow routes, certain mechanisms based on region, based on risk profile, based on any number of characteristics, and to be able to deterministically dictate that, then leverage AI where it’s applicable.”
“And making sure it fits within the risk profile,” he adds.
