Silicon Valley Bank (SVB) made headlines for its fall, leaving many bank executives to ponder what exactly happened and what they can do to protect their banks and themselves. Better managing liquidity risk is among the obvious considerations, but what are the long-term strategies?

The advancement of AI and the changes toward open banking add new dimensions and challenges to the market turbulence, systemic risks and many uncertainties. This article will provide key executives, including CEOs, CFOs and CROs, with cost-effective finance transformation strategies to complement digital transformation initiatives in-flight in most North American banks.

We’ll address the immediate concerns and discuss how we can progressively transform the finance, risk and regulatory compliance ecosystem leveraging AI, specifically machine learning and deep learning, and robotic process automation (RPA) to create new business capabilities and enforce governance for the new technologies to mitigate risks and sustain growth.

Why Finance, Risk And Regulatory?

Finance, risk and regulatory compliance are at the heart of banking areas. Still, many banks deprioritize these factors due to the complexity and scale of transformation, where there are hardly any quick wins or shortcuts. But any mistakes therein will severely impact the bank or lead to the dismissal of the executives managing that division or function.

My company’s advisory team and subject matter experts work with senior executives and delivery management daily on digital transformation in these selected domains and recommend that the banks must be prepared to root out the deep and hidden causes that surface as various forms of risks. Some of the risks can manifest themselves as short-term profitability but can eventually lead to the bank's demise.

The flip side of seeing the risks clearer is that banks must forecast better, allocate the cost with deep insights, and transfer the fund with speed and precision. This is where AI and RPA on the cloud come in handy to complement and accelerate the outcome-based transformation strategies in domains of finance, risk and regulatory compliance.

To do it right, however, governance and guide rails must be enforced. Banks are experts in banking, after all, not technology. Oftentimes, collaborating with vendors as well as regulators in a particular business or technology domain can be the most sensible approach.

Continuing The Financial Transformation And Building New Business Capability

Most banks have started digital transformation to build data lakes and migrate applications onto cloud platforms. Some banks have also attempted to transform finance and risk—for instance, by modernizing legacy systems.

As a part of a sustainable strategy, finance and risk groups must collaborate to create value for business partners as well as technology custodians and to fit into the bank’s overall priorities and vision. For instance, rather than reinventing the wheels, it is recommended to review the current state and existing programs of fund transfer, reconciliation, and cost allocation, and re-examine the target state of incorporating AI, ML and RPA to build new business capabilities—be it new forecasting models, enhanced operational processes, one source of truth for finance, risk and regulatory reporting or cloud-native infrastructure to integrate vendor solutions. The starting point can be building up or migrating existing data lakes to the federated data lake on the cloud.

One of the critical decisions is to select a cloud vendor. Oracle Cloud is one good choice, but Amazon Web Services (AWS), Azure, Google Cloud Platform or a combination of these and others can all be good choices as well. The key is to build up the virtual meta-bridge to enable federated architecture and progressive operational processes with the help of optimized caching for data, which in turn enables virtualization for consumers for timely business results.

Guardrails And Governance: Applying STAGE And TOTAL Security To Machine Learning

If you need to read more on the banking STAGE framework and TOTAL security, please follow the links. In a nutshell, cloud platforms offer much more robust security controls than on-prem counterparts. However, it is not a one-to-one shift-and-lift process to transform finance, risk and regulatory compliance to cloud-native capabilities. Cybersecurity has its new twists and turns due to the new capacities of ML on cloud platforms.

Safe and sound security infrastructure and governance framework must be in place for data technologies. This includes best-of-breed data lake or data mesh platforms such as data wrangling pipelines and MLOps, along with Spark-curated data vaults by business domains, streaming-enabled data fabrics and AI-augmented transfer learning leveraging edge computing for financial crime prevention biometrics.

From proof of concept to production, my company has worked with finance and risk teams to develop, train, deploy and fine-tune profitability forecasting applications, new models for liquidity risks, virtualized data consumptions patterns for external and regulatory reporting as well as internal and human-in-the-loop analytics, adjustments, reconciliation, audit screening and issue resolution.

More rigorous and challenging is the need to enforce security framework, especially in the case where sensitive data elements are part and parcel of the data wrangling and data enrichment. Synthetic data may be preferred, and cached or in-motion equivalents may be mandated. Where at-rest data are required, encryption and additional security controls must be strictly compliant with internal data governance, privacy and security policies and, more importantly, relevant regulatory requirements.

Conclusion

This article is an executive-level strategic proposal toward digital transformation by embracing AI and RPA on the cloud for finance, risk and regulatory compliance in large banks. Importantly, it only scrapes the surface and glances over some of the most important aspects of security, governance, new capabilities building and cloud-native features—which are the most challenging aspects to people who are living in the banking domains that many outsiders may take for granted.

These technology and business areas are extremely challenging. These are deep waters where sharks are swimming and potent forces emerge to significantly change, if not redefine, banking as we know it today. To start at the heart of banking and the most prominent suite of AI machine learning is not an easy task. It takes courage and talent to drive through the next decades of digital transformation.

LeackStat 2023