Digital transformation of large organisations like banks is bound to be arduous, protracted and varied. Although there may not be a panacea for this varied and complex problem, but there are a few common threads that can be discerned. This article attempts to sketch a brief outline of these common threads that banks can consider as they embark on their digital transformation journey.
Rationalisation of multiple technologies – “Brevity is also the soul of wit” – as well as of clarity
This is somewhat obvious but needs to be reiterated to challenge the status quo of existing technology. Banks began computerisation many decades ago to primarily adopt core banking systems, which have served their purpose quite well in the traditional model of bank branches and ATMs. However, banks may now find themselves entangled to many applications which may no longer be fit-to serve scalability and “always” availability demands of digital channels.
Moreover, many times, banks adapted different applications and technology stacks to serve a common business goal because of an absence of a consistent technology strategy. It is not uncommon to find that many different software functions to return a simple query like customer balance. Therefore, the first step towards digital adoption is to look into the set of applications that were not developed with a consistent digital technology strategy in mind, but were rather built to fulfil some past requirements drawn in isolation.
A consolidation of these applications will not only reduce time to market for new features, but also the cost to run the applications necessary of day to day operations.
From monolith to loosely coupled architecture – “the elephant in the room” – let’s talk about that
A loosely coupled architecture of software application enables rapid software development, increases resiliency and fault tolerance, and makes software systems easily scalable with the increase in user base in digital channels. These are also tailor-made to be deployed to cloud and utilise benefits of distributed system.
However, the software applications that sit at the back of mobile applications or website may not have been developed with these principles and are mostly monolithic, which are difficult to change and maintain. Any digital transformation can consider evolving to a loosely coupled architecture of these bank-end applications to take advantages of distributed systems.
Increased adoption of test automation – “to err is human”- so automate
Automation in tests during build to ensure quality up-fronting the development process will lead to early identification of potential bugs in the software and ensure quality up-front in the delivery process. Advancement of technology and adoption of test engineering practices like Behaviour Driven Development means that user requirement, in the form of user stories, can be written in a manner that it can be automatically converted to test cases, which will reduce manual testing efforts and will enable traceability of requirements to test cases and actual testing performed. These automatic and repeatable test cases for the entire applications in a regression test suite can enable rapid and frequent deployment of software in production without impacting quality adversely.
Continuous integration and continuous deployment – “the more the merrier” – let’s deploy often then
Automation of build pipeline of software deployment including automated code scans not only for coding standards, but also for any potential security issues, performance bottlenecks integrated with automated test cases can create software which is always release-ready. This automated and accelerated product release processes will lead to on-time and frequent product releases reducing or eliminating system down-time completely. Moreover, production release should eventually become a non-event with little or no manual effort needed from engineering teams.
Conversion data points generated from customers footprint to digital channels into actionable insights–“You see, but you do not observe” – let’s do both
Right from gaining an insight into user pain-points while using an App that might lead to app abandonment, potential areas of opportunities for optimising customer journeys, to determining an offer and a deal most appropriate to the customers, advances in machine learning algorithms can offer a lot features to be developed for the benefit of customers.
Data and resultant insight into app usage pattern of customers can also make it possible to provide contextual help to the customers as they are using the app, find out patterns in their transactions leading to better detection and prevention of fraudulent usage. It can open up possibilities of almost a separately channel of interaction for the customer, where they can interact with the app using audio-visual inputs instead of current touch or keypad-based textual inputs.
Real-time visual monitoring, logging and altering systems– “nothing can dim the light that shines from within” – let system logs shine a light upon themselves
Logging and monitoring systems can now deliver actionable insights in real time from centralising infrastructure logs and metrics to keep a pulse on customers concurrently using the app or websites. The response time to server requests from digital channels during a certain time-frame can be monitored for an average, or a maximum or minimum response time.
If a request is detected to be taking too long to process beyond normal limits, an alert can be generated automatically for further investigation. Similarly, system load over a set timeframe can be monitored which is useful for traffic tracking. If there is an unusual spike in user activity, e.g. in non-business hours or on weekends, this can be altered and investigated. It could be caused, for instance, by web crawlers who index the website content or evil bots scanning system for vulnerability.
Many of the monitoring tools have a built-in alerting engine (e.g. email or Slack notifications) for some conditional rules. This makes it possible to move away from a reactive system monitoring to a proactive and aggressive approach to resolve system issues.
This also helps in reducing manual efforts to operate a production system and time in debugging the underlying cause for an issue while providing a superior customer experience by increasing resiliency and fault-tolerance and availability of the system.
(Views expressed in this article are a personal opinion of Pinak Chakraborty, Senior Vice President of Technology, Digibank at DBS. He is developing a mobile first and paperless banking experience for customers.)