Getting Intelligent Automation Right in Financial Services

One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man.

- Elbert Hubbard (Writer)

Getting Intelligent Automation Right in Financial Services

The world today is complex in terms of processes, technological advances, and the markets driving them. While automation of routine manual tasks has become easily attainable, there is an immediate need for infusing cognitive intelligence into business functions in a creative and innovative manner.

Technology has been advantageous for elevating human productivity overall. However, it also presents many operational considerations and challenges like having to manage data quality and complexity, future-proof the technology stack, standardize processes, and enable flexibility and tractability in the application portfolios of the firm. Technology endeavors present cost challenges – hence the motivation for a cost-conscious firm to see the underlying process optimization as a Tier2 priority.

Organizations are often split over their automation strategy - what should they automate and how much? While robots are successfully taking over repetitive, voluminous tasks from humans, can they offer the cognitive intelligence provided by humans?

To search for plausible answers to this dilemma, let’s consider the financial services landscape as an example.

Present-day challenges for banking and financial services

For decades, financial services firms have been grappling with obsolete legacy systems, big data, and monolithic infrastructure. The kind of availability and extensibility of these systems that is required by users is often hard to achieve and results in millions of manual processes. Challenges due to rising operational costs, technology disruption by smaller and boutique start-ups, an evolving regulatory landscape and financial risk are becoming overwhelming, whereas firms are frequently behind in their risk assessment techniques. One crucial, yet unmitigated risk is cyber fraud. These challenges will continue into the conceivable future.

The unanimous solution to this is “modernization”, and its scope is massive. While it is not always viable to re-architect systems from scratch, financial services firms are spending heavily on technology overhaul and infrastructure upgrades. They are increasingly adopting solutions that encompass data scraping, data entry, and data validation combined with optical recognition technology to eliminate manual tasks. Aside from selecting the right tool for the application, they must look for service, analytics, workflow, and digitization capabilities as well as Machine Learning (ML) with an aim to reduce manual controls. In such cases, choosing wisely is essential to quick and lasting benefits.

Large financial services firms have set up automation centers of excellence with required infrastructure and defined their automation frameworks and toolsets. Despite significant success with their initial use cases, they still experience setbacks (described below) during the critical stages of production that can result in an opportunity loss.

  • End-to-end assessment is not done prior to starting the initiative(s).
  • Even if assessments have been done, there is no solid roadmap that aligns with the business vision.
  • Incomplete information is offered by operations personnel (subject matter experts).
  • There is “Bottom-up” opportunity assessment i.e. “use-case shopping”.
  • Lack of detailed process documentation and limited bandwidth of subject matter experts could lead to delays in discovery.
  • Conflicts of interest might occur with third-party BPO resources.
  • Fragmented processes might result in sub-optimal opportunities.

As firms embark on this transition, they must partner with the right solution provider for designing the journey from start to finish. In addition, bringing in a third-party consulting service will unquestionably deliver additional benefits. The objective third-party can perform a holistic process assessment, enable selection of the right solutions, and deliver the right metrics and performance in line with their business goals. Infosys Consulting, with deep knowledge and vendor partnerships in this area can provide the governance, organizational communication, and change management that is essential to a successful transformation.

Benefiting from a structured assessment framework for intelligent automation (IA)

The discretionary artificial intelligence (AI) and automation projects and resources pivots towards easy, quick wins (low hanging fruit). Take for instance, Robotic Process Automation (RPA). Even though RPA is helping drive quick results, the real intelligence needed for fully automating business processes is not always being prioritized.

In contrast, IA comes with RPA capabilities in addition to being able to process automation enabled by bots that learn from data and adapt in real-time. A third-party assessment must, therefore, focus on quick gains without losing perspective on the long-term shift towards innovative cognitive intelligence that can be gleaned from IA. They will need to perform a deep study of each case at hand and identify opportunities for today, tomorrow, and the day after. The assessment process is followed by selecting the best vendor. It must quantify immediate gains and enable long-term value. Such structured assessments enable scaled automation, thus eliminating the risk of acute “value leakage,” which may happen when the actual value realized falls short of the expected value.

To do so, it is important to have a clear design and definition of the solution for changing business requirements and processes – one must be able to trust and estimate the outcomes of IA.

At Infosys, our assessments enable specific cost take-out opportunities and deliver value in the long term. These are shown in the visual below.

Our structured assessment framework is aimed at finding value and design and is meticulously designed to address issues that could lead to value leakage. Once we perform the assessment, it provides a repeatable framework for forward engineering.

Risk recognition is strategic to assessment

We continuously revisit and input our findings into assessment results to avoid potential problems related to changing business requirements and processes. Our assessments identify and address key risks to successful implementation such as:

  • The process can go awry in the absence of overall control and monitoring. The design must consider and visualize scenarios to determine whether we can truly eliminate human oversight/ human intervention in a control process.
  • Today, data is everywhere, and we rely on data analysis to inform us, more than ever. However, the analytics process may not look for false positives and, thus create a high cost for the bank/financial services firm. Hence, easy ways to eliminate false positives must be considered via restructuring of the process.
  • Simple configuration errors could result in undetected, but big, irreversible losses. For example, when we automate something as complex as a loan or securities underwriting process, negative consequences can be devastating. However, since the technology is designed by humans, it is prudent to assume that it cannot be bug/error free.
  • Hitches can occur when the machine encounters exceptions and cannot handle these because it doesn’t know outside the business rules. This will be significant where humans handle decision making and complex data analysis is involved; for instance, things that are situational and need presence of mind.
  • In case the machine learns incorrectly, there can be a cost overhead to undo the learning. The design must, therefore, incorporate the right “learning” context.

Next steps

Experts dealing with product and compliance, UI and analytics, data architecture and governance must draw a detailed implementation roadmap, and plan for incremental releases. Here, the agile approach offers the best methodology as it can manage small, incremental releases that give long term benefits. Infosys can step in with roadmaps and eventually with product management, once the initial assessment stage is complete.


A proven methodology and consulting framework can uncover several automation opportunities in the financial services sector. The biggest advantage is that ours has been vetted by many clients in the financial services industry. At Infosys, our assessments are a practical guide to the best-case realization of automation opportunities today and for the future.


Intelligent Automation: A Beacon of Certainty in a World of Uncertainty
Benefits of Intelligent Automation: IA vs RPA


Kavita Sastry
Principal Consultant, Infosys Consulting
Kavita has over 20 years of experience within in-scope delivery of complex technology and operations as well as business-focused solutions for global capital markets. She has enabled the path to value realization for various financial services clients. She has worked extensively with leading banks and capital markets firms executing complex transformations in business, technology, and regulatory change management.