Overview
The current Discai product offering is focused on qualitative AML alert creation and includes both a unique set of AI models and a fully integrated rule&simulation engine. Both modules seamlessly integrate with best-of-breed third party or (customer) home-built solutions, allowing a modular and incremental integration approach.
Key Features
AI-based model suite
The AI models allow banks to fight money laundering with future-proof technology and trusted algorithms. The models enable the bank to find real and complex AML risk, which is sometimes difficult to capture in a rule-based approach. We offer a unique set of supervised, unsupervised, and graph learning models, tailored to capture the AML risks for the individual bank. All models come with a full explainability on why the models conclude that the customer is considered high risk for AML. As part of our data-science-as-a-service approach, all models are built and maintained by Discai, keeping data science resources on customer side available for other priorities. Note that the full solution has been discussed with local regulators and is actively used by several large banks in Europe.
We offer several types of models:
·A foundation model identifies behavior that is similar to previous (known) cases of potential money laundering.
·Several focus models find specific types of money laundering that deserve attention in the specific context of the bank.
·The anomaly model identifies behavior that is not in line with what would be expected, given the customer’s profile.
·The network model connects suspicious customers to each other, based on transaction behavior. The models elevate customers scores that are closely connected to other suspicious customers or (known) potential money launderers.
Rule&simulation engine
Despite the strong performance of the AI-based model suite, we do realize that some AML risks are better handled in a rule-based approach, either due to explicit regulation or due to ‘zero tolerance’ risks defined by the bank. Moreover, a bank might want to use the AI-based models for prioritization of rule-based alerts, rather than stand-alone.
The rule&simulation engine is built by compliance experts and allows the flexibility to define, maintain, simulate, and process AML detection rules in a low-code approach. As a unique feature, it allows the user simulate the impact of changes based on historical data, so that the impact of the change is immediately visible and can even be sent for investigation (our unique ‘lookback’ functionality).
Key Benefits
As no other provider, we understand and feel the challenges of a bank. Together with best-of-breed partners, Discai offer a fully integrated and end-to-end solution that allows financial institutions to:
- oMeet all regulatory expectations of a bank, in a solution that is continuously kept up to date.
- oReduce pressure on operational teams by reducing false positives and optimizing compliance processes.
- oUse advanced and future proof technology to fight financial criminals with equal arms, while ensuring responsible use of technology.
- oPrevent fines and negative impact on reputation, fit-and-proper assessments, and adverse selection.
Our solution and its implementation approach are designed to maximize real business value with minimum risk:
- oWe offer not just technology but a data-science-as-a-service proposition, combined with industry best practices on how to use technology in a highly regulated environment.
- oIt was thoroughly discussed with European and local regulators and actively used in several (large) financial institutions in Europe.
- oThe modular architecture and incremental implementation approach limits the impact of ICT teams and the organization in general.
- oWe aim for a partnership approach with easy accessibility in terms of product improvement, regulatory alignment, training, incident management, and management alignment.