Subhasis Bandyopadhyay, General Manager, Mindtree discusses the pressures of making regulatory decisions at scale and how machine learning can help…
Banking and Financial Institutions (FIs) are no strangers to compliance and regulatory processes. In fact, with GDPR (General Data Protection Regulation) recently coming into effect this year, I’d hazard a guess that every business, regardless of the industry in which it operates, has likely had to brush up on even the most basic aspects of its legal Governance, risk management and compliance (GRC) strategy.
Increasing complexity and changing demands
Today’s landscape within financial compliance and regulatory risk is one of increasing fluidity and complexity, thus increasing the potential for older, rule-based models to fail. In order to prevent financial and reputational loss, these firms need to move beyond a static, rule-based system and adopt a new approach if they’re to keep regulators at bay.
It has become apparent that given the expanding compliance expectations, current check-the-box rule-based systems do not provide adequate coverage and as such, operational efficiency is becoming critical in current banking applications. Banks and FIs are facing increased risks with the rising sophistication of financial crimes as existing systems only address a known set of scenarios, whilst finance and product controllers are being forced to adhere to strict operating and reporting standards in the new regulatory environment.
What’s more, to keep on top of all this whilst scaling appropriately is a task unto its own. Raise your hand if your organisation has to deal with a high number of false alerts, extensive time-consuming audits, labour-intensive investigations and error-prone processes on a somewhat regular basis. Yes? Thought so.
Making smarter decisions
Given the demand and subsequent expenditure of these practices, banks and FIs are increasingly turning to automation. Businesses are turning to those offering AI and machine learning powered predictive analytics tools that use self-learning mechanisms to continuously evolve with new data points and user analysis. This allows for a much more dynamically adaptive system that produces better prediction accuracy and drives operational efficiency. For example, Mindtree’s recent partnership with Tookitaki brings together Tookitaki’s Decision Support System (DSS) built on AI and Mindtree’s offerings in the banking and financial services domain to provide solutions to help businesses make better decisions.
Businesses require a dynamically adaptive model for alert management based on AI and machine learning that is relevant to business at all times. The system must be an automated, highly accurate, self-learning solution that detects new, suspicious cases with ease. Additionally, the ideal alert management solution should also offer end-to-end integrated services with advanced analytics for improved business agility. Existing rule-based alert systems fall short of requirement as they address only a known set of scenarios. As a result, businesses are likely to see a high percentage of false positives and negatives while missing out on the truly suspicious ones. This further leads to revenue loss, heavy penalties for non-compliance and loss of brand reputation and image. To solve this, Banks and FIs need an alert management solution that not only provides accurate information from multiple sources but also offers efficient fraud identification and management in real time.
As well as a dynamically adaptive alert management system, businesses must improve their reconciliation management system to ensure an end-to-end, automated solution to reconciliation management across workflows. This solution should leverage business processes and predict matching and exception resolutions without human intervention. The new regulatory environment is forcing finance and product controllers to adhere to strict operating and reporting standards. Given the fragmented reconciliation processes, being able to automatically handle exceptions and correct source systems is very difficult. By leveraging AI and machine learning, these solutions are able to optimise business processes and predict matching and exception resolutions without human intervention. Businesses should also aim to use advanced analytics solutions to further help transform data into insights to drive business outcomes.
The end result of applying these solutions is that businesses will see many benefits, including improved process efficiency, built-in self-learning capabilities providing data revisions resulting in updated information at all times, and reduced people dependency.
Benefits of good decision making
Through our partnership with Tookitaki, we’ve been able to help various banks and FIs to save millions in alert management and reconciliation, whilst also increasing screening accuracy and overall efficiency.
No doubt, the Governance, risk management and ccompliance space will continue to evolve into ever more complex systems over the coming years, but with machine learning and AI on hand to adapt appropriately, businesses will be best placed to keep the regulators happy and in turn make better regulatory decisions at scale.