Enterprises have been struggling to develop applications that are agile and quick to change. Microservices architecture provides a way to overcome this challenge and has therefore caught the attention of enterprise IT teams.

From a technical angle, microservices avail service-orientation principles and functional partitioning to “monolithic” applications so that the systems work as relatively independent services. From a business benefit point of view, it makes companies digital ready and provides numerous advantages to them like allowing experimentation and innovation, sanctioning more autonomy and accountability, enhancing the resiliency of mission-critical systems, by scaling and reproducing parts of systems rapidly, and enabling cloud adoption and cloud-native architectures by deployment of services over multiple VMs or containers that function together.

Bank regulators are looking to drive better deals for the clients by generating more competition by innovating customer information sharing, transaction initiation, and payment mechanisms through open APIs. With an open API economy, banks get an option to introduce their customers to new products and services through collation among the business units within a bank, across industries, and between banks and other complementary sectors of the economy, especially data businesses and technology.

Microservice splits monolithic applications into a set of services that communicate with each other through open APIs. Each service acts out one particular function pretty well; the service and its API are a product that is detectable, well-defined and carefully maintained. These self-contained services are then constituted as needed so as to deliver complex functionality, even if they are deployed independently of each other. Services can calibrate independently as well, making the software flexible at runtime. And if a single service folds, it doesn’t mean it will bring down the entire system because of the flexibility built into microservice-based digital banking solutions.

It has come as a blessing for the banks because this approach will provide favorable circumstances to deliver value at subsequent intervals and not wait for a massive technology shift to happen. Open API banking combined with microservices-based architecture will define the success in banking space in the near future, where banks would be in a better position to accelerate time to market in the real sense.

While microservices promise a better future to the banking industry, it would require banks to rethink their operational support, delivery methodology and required skills. Only that banks that will consider an upgrade in people, process and technology will be able to leverage the power of microservices and utilize it to the maximum.

Examples of banks that have benefited from adopting microservices

Nordic APIs recently worked with the Canadian Imperial Bank of Commerce (CIBC), who have set out on their own microservices journey to get rid of their existing monolithic backends. One crucial tenet of this journey is a microservices framework Light 4J, that CIBC built themselves. They’ve assembled it from open source components instead of trying to customize a commercial offering. Nordic APIs was assigned the task of reviewing its features and functionality to both validate their progress and provide constructive feedback.

By developing their own microservices framework, CIBC acknowledged the need for microservices in a standard, reiterating framework of development that can be embraced by the all within an organization. Cultivating a microservices framework from the beginning connotes that it can develop side-by-side with CIBC’s microservices maturity. This is crucial because it allows the organization to learn lessons and adjust their approach accordingly. With these benefits, CIBC is well-placed to adapt and experiment with their architecture, both in terms of domain and technology. This kind of flexibility lends itself to the Evolutionary Architecture style. In the process of building the Light 4J framework, CIBC is using a standardized microservices frame that deals with cross-cutting concerns as a key tenet of the framework. The homemade approach is also beneficial when it comes to introducing resiliency in the style of API that can be supported.

Another example is the Monzo Bank, who created their backend systems using microservices architecture and implemented using Google’s Golang. The memorable part of this project was Golang’s brilliant concurrency primitives that make this language consistent for creating ‘high volume, low latency, distributed applications; the use of a microservices framework like Monzo’s open source ‘Typhon’ framework in amalgamation with the CNCF ‘linkerd’ proxy, being highly advantageous for implementing core communication concerns; and enabling distributed tracing through context propagation being a key enabler for observability and debugging distributed systems.

Will the microservice architecture become the preferred style of developers in future? It is a question that still remains unanswered. But it’s clearly a potent idea that provides serious benefits for designing and implementing enterprise applications.

In the recent times, microservice architecture, or simply microservices, has become quite popular. To put it into simple terms, microservices is a characteristic method of developing software systems and applications as a suite of independently deployable, small, modular services in which each service runs a distinct process and communicates through a well-defined, lightweight mechanism to serve a business purpose.

Origin of microservices

Microservices has a profound historical background. The term ‘microservices’ was used for the first time in mid-2011 at a workshop of software architects. After James Lewis laid out  his ideas about microservices in 2012, several IT companies started considering the idea of microservices, and by 2014, large enterprises had started serious discussions about investing in it.

In 2014, James Lewis and Martin Fowler collaborated to provide a few real-world examples and presented microservices to the world. The major tenet of a microservice is its ability to take the place of services in a system that function independently from each other, the configuration around business capabilities and using smart endpoints and dumb pipes.

In the original object orientation model from the 1970s, an object is like a little computer that you send messages to and program it to perform tasks. The actor model is established on similar concepts, where the actor is the computer to which you send messages through its mailbox. Both of them are antecedents to the core aspects of microservices.

The idea came from the need of isolation between services. When you run every service in its own process within the same server, it is assured that a contract is followed. If all microservices are run inside the same process, all services will be eliminated if the process is resumed. When each service in run in its own process, it is ensured that only one service is eliminated in case the process is reinstated but if the server is restarted, it will eliminate all the services.

SOA vs Microservices

While many confuse microservices to be another name for Service-Oriented Architecture (SOA), they don’t know that traditional SOA is a broader framework and can mean a wide variety of things. Although they both rely on services as the main component, they vary greatly in terms of service characteristics.

The characteristic SOA model has, for example, more dependent ESBs, with microservices that use faster messaging mechanisms.  SOA also emphasizes on imperative programming, whereas microservices architecture focuses on a responsive-actor programming style.

Additionally, SOA models generally possess an outsized relational database, while microservices usually utilize NoSQL or micro-SQL databases.  However, the real variation lies in the architecture methods used to arrive at an integrated set of services in the first place. Also, while SOA places more importance on business functionality reuse, microservices emphasize on the concept of “bounded context”.

Benefits of SOA and microservices over monolithic architecture

Before we discuss the benefits of SOA and microservices over monolithic architecture, it is important to understand what monolithic architecture is. Monolithic architectures work on a single application layer that brings together all the functionalities required by the architecture. It is the simplest form of architecture because it doesn’t involve as many actors as other architectural styles.

The monolithic model works perfectly for small and mid-sized architectures and keeps the complications quite low: the problems usually appear when the architecture needs to scale up feature-wise since the modules are extensively dependent upon each other. SOAs usually incorporate functions into small/mid-sized applications and try to keep the complexity of each app / functionality very low and make them communicate over a set APIs (being them HTTP APIs, asynchronous messaging and so on).

In contrast, microservice capabilities are expressed formally with business-oriented APIs. They compress a root business ability, which are prized assets to the business. The execution of the service, which might require integrations with systems of record, is completely concealed. Dependencies between services and their consumer are diminished by applying the principle of loose coupling. By standardizing on contracts expressed through business-oriented APIs, consumers are not affected by changes in the implementation of the service. This allows service owners to change the implementation and modify the systems of record or service compositions which may lie behind the interface and replace them without any downstream impact.

When positioned properly, a microservices-based architecture or an SOA can bring substantial value to the business. That value can be expressed in both technical debt being avoided and a considerable increase in efficiency.

The Mobile First Approach to Digital Transformation

Digital transformation has evolved into a strategically imperative business operation for all industries – with mobile phones playing a key technology-enabler. In a bid to ‘mobilise’ banking, financial institutions are recognizing the role of smartphones in digital transformation and its integration into core business strategies.

In a recent report by Cisco it was stated that by 2021, there will be nearly 12 billion mobile-connected devices generating on an average of 5.7GB of mobile traffic per month, per user. While mobile banking apps are not new to the finance industry with 31% of consumers using mobile apps the most, as per Citi’s 2018 Mobile Banking Study. From checking their balances to transferring funds, apps are taking care of all banking transactions. Rapid development in mobile banking technology has shifted user’s perception towards digital payments and also seen an increase in “digital branches” that replace traditional banks. The Royal Bank of Scotland (RBS) is working on plans to create a standalone digital bank and is soon to shut down over 259 branches across UK. With 5.5 million mobile banking customers, RBS is focusing on its mobile-consumers that prefer contactless technology.

Why should you consider a mobile-first approach?

As per Deloitte’s 2018 Banking Outlook report, the emphasis is on building a mobile-centric platform that’s redefining their retail channel strategies, product portfolios, and delivery models. The report stresses on the importance of viewing mobiles, not as a one-way window for delivering products and services, but rather places it at the core of delivery strategies.

• Customer-centric approach

Consumers are constantly looking for services that are convenient, available at all times and do not compromise on security/privacy. To meet the demands of a growing tech-savvy consumer base banks have to quickly adopt technological solutions. Smartphones have become essential to deliver a seamlessly connected omnichannel customer journey. With a heavy consumer-base, an interesting trend seen in banks is the move towards design and user experience.

Mobile banking creates new rules for a customer-centric banking process – smart visual interface, building a relationship with the customer through other channels, generate traffic to increase sales and, improve speed-to-market rates with innovative solutions.

• Options for cross-selling

An effective mobile banking strategy would include channels to improve cross-selling rates. A decline in bank foot traffic is making banks develop selling strategies that rely heavily on technology. Machine learning techniques are empowering financial institutions with data insights (such as app usage, location, transaction history etc.) that banks can leverage to increase sales. Analytics is also being used to acquire new customers through digital channels.

• Breaking organizational silos

As banks transform into digital organizations a change in organizational structure is evident. Legacy systems are quickly being replaced by a unified system that operates irrespective of the technology stack evolvement, improves efficiency and is easily scalable. PSD2 has been the biggest driver for breaking down data silos, wherein banks have to start sharing data with their party providers. The new directive opens up banking industry to new entrants and is creating a shared data network system, where information is easy to access and share across all touch points. Banks, like HSBC, are adopting technology to centralize their operations and expedite services such as loan processing.

• Keeping up with technology

The increase in connected devices is an opportunity for banks to capitalize on IoT and wearables. “Wearable banking” is soon going to be the next big thing to impact banking-on-the-go. It is forecasted that the Internet of things in banking market is expected to register a CAGR of 18.49% by 2023. Banks that succeed in keeping up with technological trends will become trendsetters for creating numerous downstream opportunities.

• Competing with non-banks

With the new PSD2 directive banks have to find innovative solutions to compete not only with each other but also with non-banking organizations like Google, Amazon, PayPal etc., to retain their consumer base. For example, Apple recently announced the launch of Apple Pay – a secure payments service for iOS. It also provides loans to small and medium-sized companies. While organizations like Apple are not looking to turn into a financial institute, it is working towards partnering with banks to grow its financial stance.

• Improving enterprise mobility and employee efficiency

While external innovation is taking place with the presence of apps and wearable-tech, financial institutes also have to look into improving workforce productivity. The right mobile strategy enables the bank to adopt digital transformation as a multi-faceted tool for their employees. For the mobile workforce, organizations could look at improving collaboration and productivity through mobile workstations, real-time access to customer and product information, and foster innovation within the organization.

Digital is the new imperative that is driving the banking and finance industry across the world

Five years ago, an AT Kearney Study had said that digital is touted as a revolutionary transformation that will bring many new features, including any time and anywhere banking, ultra-fast response times and omnipresent advisors. So a natural question people asked then was, why don’t banks pick up the pace? And the answer was simply that the banking and financial institutions needed to make fundamental changes to the way they operated. Well, that was in 2013. In the five years since that report, we have seen some tremendous thaw in the industry which has completely embraced digital.

The big change is that Digital is no more seen in a silo of client experience alone. The question today is: how can digital help quicken business decisions, manage new risks effectively and adhere to compliances better. If the client experience really has to jump to a new level, institutions have realized that they should go deeper with digital and change legacy standards and protocols.

For instance, banks and financial institutions are serious about deploying Cognitive Computing to increase productivity by bringing in automation and using the power of data. Cognitive Computing is believed to bring in processes, data and analytics to the point where institutions can service customers at the speed of thought, match customer expectations with specialized products and portfolios on the fly.

To illustrate the point, we look at a couple of key developments. The UK banking system is being revamped. UK, considered to be among the most conservative countries when it comes to technology reforms and adoption in banking and finance. However, it is about to witness a sea change in its retail payment system. While UK market had opened up to digital at the client level, it had held off core banking reforms. This is set to change. The Bank of England has recently approved a new system that will bring in a new clearinghouse, a faster payment engine with new standards and protocols. The NPSO (The New Payment System Operator) in the UK is mandated with building the next generation Open Banking platform. This offers new product opportunities for developers and vendors. It also addresses security and standards requirements that the new digital calls for.

Cross-border payment platform Swift is deploying Blockchains to help corporates track their payment more precisely and pinpoint reasons for hold-ups and delays. Swift helps corporates look at the payment journey from a new viewpoint using Distributed Ledger technology giving vital insights to its corporate users. Whether they are documented components, the purpose of payment, each node in the chain certifying the transaction increases the trust in the network.

In the US, AIX, an AI-powered brokerage house is set to disrupt the decade’s old interdealer brokerage model using Cryptocurrency. AIX is close to disrupting the model where voice negotiation is key in tradings. Typically, a trader wants to execute a transaction, picks up a phone and communicates to other brokers, match prices and conducts the trade. The system needs highly skilled manpower to do it which is expensive. Now, AIX is deploying an AI engine called the Rainbow Cognitive Engine which takes trade information from traders and distributes it throughout the network and brings back relevant information to the original trader. It brings in all kinds of data points across the market through the AI engine’s built-in logic and optimizes the outcome. AIX’s compelling value proposition is that all the markets will be available for a trader on a single platform. One can trade in US bonds market and invest the proceedings into Euro currency market or into Cryptocurrencies for example.

If you check the state of digital push in various regions, we have come a long way. Historically, US, Canada, Australia, UK and central European countries were quite committed to going digital. The birth of Fintech in the Silicon Valley demonstrated the power of applying technology in the sector. Most institutions in these regions began partnering with the young fitech startups to offer value-added services. However, with the growth of data analytics and big data push, banks and financial institutions in the regions committed to Digital soon realized that unless their internal teams were aligned to these new technologies they would not be able to withstand the competitive wave blowing across the globe. This had a cascading effect as bringing new technology in-house mandated that the institutions need to change their internal processes and governance. These regions are now aggressively pursuing the goal of changing core systems to align with the Digital model.

Other countries have been little more cautious. The bolder among them were happy wetting their feet with online and digital value-added services and watch for signals from experiments in going deeper into Digital. However, a quick look at state-of-affairs in more conservative regions shows us that EMEA, South America, and Asia have all jumped on to the Digital bandwagon without reservations. India is one big example which has made a huge jump to digital with traditional banking and finance institutions partnering with fintech companies. Far away in Chile to things are changing rapidly. The San Diego-based Banco de Chile is moving very fast into digital solutions aimed at making customer experience easier. Not to be left behind the Islamic Banking players too have started offering online and digital experiences. These offerings match other solutions except they adhere to the Islamic banking rules.

It’s only a matter of time before Digital takes over the banking and financial sector. The last bastion is the Regulations and Compliance front. With Digital, new risks are emerging and the authorities are racing against time to promulgate new policies, standards, and protocols. If you see Basel 3 Regulations, be it the need to keep data protected, keeping a trail of all electronic transactions or providing complete information on the website for making the informed decision, it is amply clear on how Digital is gaining importance.

References:

Interview of the following people by TheBanker:

• Likhit Wagle, IBM
• Paul Horlock, Chief Executive Officer, New Payment System Operator (NPSO)
• Tom Halpin, global head of HSBC’s payment products
• Steve Compton of AIX

The AT Kearney Report: Banking in a digital world

IFF & McKinsey: The Future of Risk Management in the Digital Era

As it has always been, today too, technology managers have to contend with two complementing technologies for their attention and budget. In this circumstance, let us review where the tilt should be.

Let us circumscribe our scope to the BFSI vertical. The principles derived here, with some or no modification, may find application elsewhere. After the 2008 financial crisis, this vertical has been facing a global phenomenon of slow growth and immense pressure on cost management. This is directly forcing technology functions to deliver results that, in part at least, would give the leaders a breather, if not a differentiator. The review is made in such a context.

Revenue vs. Cost

There are some banks that have started their effort in rationalizing their customer and product profile as a precedent to focusing on other elements of cost. If these are pruned, they can, on their own, lead to economies of operations. There are banks which have got rationalization of upto 60% on their customers and products and are still not seeing a major impact on revenue; their costs are showing signs of decline or have declined a bit, and are on their course to further decline.

Most of these institutions have used AI which comprises of machine learning, natural language processing, algorithms and workflows to achieve their result. Unfortunately, blockchain is not one of them and this is not an intended domain for that either.

Width of coverage

There are products in financial services which have varying degree of dependencies on external parties for their fulfillment. Most of the investment or wealth management products have a large component on their interface connected to markets and counter parties. Same is the case with trade finance. In all these, while there are need for external interfaces, there is equal if not more spend on the internal processes for fulfillment.

Blockchain essentially has a play where there is external dependency which is critical to the completion of the process. AI has application irrespective of this.

Internal vs external dependency

Blockchain requires a substantial consortium of banks to institute common ground to transact business among themselves and to establish the process for bilateral or multilateral transactions. Its economies are derived from such common understanding.

AI is a technology enabler which an institution can apply irrespective of the need to strike such common grounds with any of their trade partners.

Maturity of the technology

Proven platforms have begun to emerge in financial services. Some of them manage risk and adaptation of product to suit the customer while a few of them also provide decision aids about the customer for closure of the transaction. Kabbage, LendUp and Lending Club are examples of the platforms.

In addition, there are tools that are available which have focused either on a process, like customer acquisition or retention or for a vertical, say financial services, and have established their maturity. Dataminr and Alphasense on sourcing data across various platforms and also types whether text, audio, visual or graphic are a case in example. Similarly, Blue Prism and UI Path have established that they can bring acceleration to process automation irrespective of whether they relate to front, mid or back office functions.

It is hard to point out any platforms or tools that have comes to similar maturity on the Blockchain front. It is not to state they are not there. Trading platforms like Hedgy, LedgerX and TeraExchange as well as issuance platforms like Linq or 10 are a few examples. They are, however, not comparable in terms of maturity to their AI counterparts.

Adoption and practice

There are two broad trends that are emerging in AI  – Fintechs are coming out with platforms that are robust for adoption by even bigger banks; large banks themselves are aggressively pursuing their effort to streamline and economise their cost through adopting and adapting these tools. A few small and regional banks too have started their efforts along the same lines.

As regards Blockchain, consortium of banks like those in Australia or among large investment banking institutions in US have started their effort to derive value out of this. We need to see where we head from here.

Conclusion

At this time, both in terms of versatility and adoption, AI is ahead of Blockchain among the financial services vertical. What the future holds, one needs to wait and watch.

*This article was previously published in Finextra on May 9, 2017