By K R Venkatraman, VP & Head – Product Architecture, Infosys Finacle
India’s financial services sector is growing at an unprecedented pace. In July, UPI-based instant digital payments once again broke the record by crossing 19.4 billion transactions. eCommerce is growing rapidly towards a $200 billion market size by 2026. The Open Network for Digital Commerce (ONDC), Open Credit Enablement Network (OCEN) and Unified Lending Interface (ULI) are promising to revolutionise digital lending, particularly for rural and small business customers.
Even the Central Bank Digital Currency (CBDC) pilot, though still in pilot mode, is pushing the boundaries of how money is issued and used. These developments are not just urban-centric — they’re reaching rural areas and small businesses, creating new opportunities for financial inclusion. From a technological standpoint, how should banks deploy their resources for maximum scalability and optimal total cost of ownership (TCO).
As banking moves beyond traditional branches into apps, platforms, and third-party services (like embedded finance and marketplace lending), banks must handle millions of small-value transactions daily. This requires systems that are not just fast, but scalable and resilient.
Indian banks are leading the way in core modernization, outpacing many of their global peers in adopting next-generation technologies and digital infrastructure. Yet, to sustain this momentum and remain competitive in a rapidly evolving financial landscape, further investments are essential. A senior executive from a leading consulting firm estimates that Indian banks may need to invest $1 billion over the next 5–10 years to modernize their core systems.
How Banks Can Scale Smarter
1. Make applications perform and scale better
The Scale Cube model is a useful way to evaluate the expectations and performance of modern-day applications. It describes three dimensions of scale:
- The x-axis is horizontal duplication – running several identical copies of an application to handle a bigger workload, like adding more counters at a busy bank branch. Horizontal scalability is quick and easy to implement and improves fault tolerance and system reliability. This is the traditional way of scaling.
- The y-axis is functional decomposition – achieving scalability by decomposing a large application into specialised functions that are scaled independently, for example separate systems for payments, loans, and customer service. Leveraging cloud and microservices architecture, it helps small teams manage development more efficiently. This is where cloud native architecture and microservices come into play. With capabilities like Horizontal Pod autoscaling and pay per use models, cloud native capabilities help optimize TCO. Right grained microservices, that are aligned to a domain bounded context, help with the functional scale.
- The z-axis is about data partitioning – where data is partitioned, sharded or entire workloads are carved out to be processed on alternate, light weight processing engines. This paradigm takes computing into the eventual consistency model and banks have to make a risk vs consistency decision with this model.
Together, these approaches help banks build systems that can grow with demand. For instance, during festive seasons or salary days, banks can handle spikes in UPI or IMPS transactions without slowing down.
Example: A bank might run multiple payment gateways (X-axis), separate its loan processing from customer onboarding (Y-axis), and offload UPI processing to a purpose built, lightweight UPI processor.
Applying the scale cube framework is one element of modernisation; by making applications scalable and performant, it takes banks a step closer to legacy core transformation. Of the various legacy core modernisation options, “hollowing” is one of the most popular choices.
2. Progressively divest the core’s capabilities
Many banks are adopting a strategy called “hollowing out the core.” Instead of replacing the entire core system at once, they gradually move specific functions like customer data management, payment processing, or loan origination to newer, more efficient platforms.
By carving out high-volume, high-performance workloads – UPI payments processing is a good example – from their existing cores, banks can make their legacy platforms more manageable, reduce the (typically high) total cost of ownership, and scale the migrated workloads with ease. Over time, the new, surrounding components take over the bulk of processing from the legacy core systems.
3. Migrate to private, then public, cloud
Cloud is without doubt the most important technology in core modernisation. Shifting existing workloads to a modern core built on cloud-native architecture yields major benefits, including scalability, resilience, agility, ease of innovation, and cost reduction.
A smart approach is to start with a private cloud, where banks host systems on their own secure infrastructure. This would not only enable them to reap the benefits of cloud within the security of their own premises, but also allow them to build confidence and experience before the subsequent move to public cloud.
Summing up
Digital public infrastructure, such as UPI, OCEN and ONDC, and changing customer behaviour are causing India’s banking transactions to grow exponentially. While new-age digital providers are cashing in on the opportunity, incumbent financial institutions hamstrung by their legacy systems are struggling to meet market needs. In order to remain competitive, Indian banks should urgently modernise their technology estates into scalable, agile and resilient systems that can support rising transaction volumes with ease. The scale cube model, core hollowing, and private cloud play a key role in this transformation.