A potential compliance issue was expected to require an additional 950 FTE in Operations to solve. The CEO of the bank asked Technology to look for alternative AI and Machine Learning based solutions to reduce the additional FTE required in Operations to achieve compliance.

The CTO asked me to run a cloud based AI and Machine Learning Proof of Concept to prove that false positives from sanctions screening could be reduced to an operationally feasible level and minimise the additional FTE requirement in Operations.

Mobilised a team of business analysts and architects to run a PoC to determine the feasibility of using an AI & Machine Learning solution instead of increasing operational headcount by 950 FTE while addresses the potential Sanctions compliance issue.
Conducted market analysis and extensive vendor assessments to create the Business Case and projected ROI for the AI & Machine Learning solution and associated operational efficiencies.
The team created a target state operating model to support the wider Digital Transformation and understand the data, process and people changes required to support a complaint cloud-based solution.
Commissioned an AWS IT Security Architecture and led the Data Governance work to safely experiment with production data in the Cloud.

The PoC proved that compliance could be achieved with a substantially smaller increase in FTE (90% fewer than initially estimated). This resulted in a >$5m saving for the bank leveraging AI and Machine Learning technology to strengthen its Financial Crime compliance. The PoC was deemed successful and was operationalised into production. The solution went on to uncover additional Financial Crime risks that were previously undetected while freeing up Financial Crime analysts time, from low value tasks of clearing false positives, to higher value tasks such as reviewing exceptions and regulatory RFIs.