To improve the detection of financial crime without compromising sensitive information, our participant Knights Analytics has developed a solution. This utilises AI and can be combined with Multi-Party Computation (MPC) to link records about the same real-world entity across multiple banks, thereby creating a combined view whilst minimising sensitive and/or confidential information being shared.
The creation of this combined view by linking records is called Entity Resolution. Entity Resolution can be conducted by either deploying relatively simple rule-based algorithms, or more sophisticated AI algorithms. Knight Analytics leverages both techniques to solve complex data challenges in organisations where data is messy, disparate, and stored across multiple siloes. Creating a combined view of clients and their transactions by collaborating on data can help financial institutions to combat financial crime because it provides them with more complete insights on actual transaction patterns and organisations.
The Centre of Excellence for Data Sharing and Cloud (CoE-DSC) has supported Knights Analytics to assess how the introduction of MPC influences the performance of Entity Resolution in terms of quality and scalability and for which other use cases the developed solution could be utilised.