How Intelligent Automation of AML Will Benefit You | WorkFusion

How Intelligent Automation of AML Will Benefit You

Highlights from the Capgemini & WorkFusion webinar on anti–money laundering solutions

By: Mark Fischer, Senior Director, Intelligent Automation, Capgemini; and Theo Piletski, Global Head of Intelligent Automation Practice for Financial Services, Capgemini

WorkFusion and partner Capgemini joined forces to highlight the latest automation efforts in “Know Your Customer” (KYC) and fighting financial crime in a recent webinar, available on demand here.

Capgemini encourages clients to start their automation journey with Anti-Money Laundering (AML) processes, as these will quickly improve speed, quality, and traceability. Financial institutions’ spending on automation, specifically intelligent automation of AML, is going up, thanks to increasing compliance needs for fraud detection processes, software tools, people and losses due to regulatory fines, and revenue loss due to fraud. AML processes are highly regulated; hence they are standard across the financial services industry.

To make this first step easier for our customers, we’ve invested into building a suite of bots specifically for AML. They are pre-built on WorkFusion’s machine learning-enabled Intelligent Automation platform. Additionally, we’ve pre-trained all models on 3rd-party data, to accelerate on-premises deployment and realization of automation benefits. Some AML areas we’ve covered with our solutions so far include use cases known as Adverse Media Screening, Ultimate Beneficial Owners and Sanction Screening.

Current state of AML efforts

We estimate that, on average, there are 3–5 full-time equivalents (FTEs) dedicated to each area in small- and mid-sized banks, a number that is up to 10 times greater (30–50 FTEs) in larger banks. In addition to obvious FTE savings, our solution also helps address the following challenges within the current AML software ecosystem:

  • 90% of all alerts from automated searches are false positives, which means tremendous wasted efforts from Ops teams
  • Only 4% Suspicious Activity Reports result from automated alerts
  • 2 hours is the average time an analyst spends on each investigation for small- to mid-size enterprise customers

Benefits from Intelligent Automation of AML


  • Companies can fund this work with the savings achieved by selecting and prioritizing AML processes with the highest ROI. Because these solutions are pre-built, the cost is low, but the impact is high.
  • Business can grow without increasing the AML workforce and its expenses.
  • Cost savings also are realized by avoiding hefty fines and loss of revenue


  • Demonstrating industry thought-leadership and recognition.
  • Helping to do the right thing effectively and efficiently; prevent money laundering and terrorist financing, market manipulation, and other offenses
  • Mitigating risk of reputational damage

Highlights from the webinar discussion

How does the Capgemini solution differ from WorkFusion’s packaged Adverse Media Screening /KYC business process?
The Capgemini solutions and WorkFusion solutions work together, weaving different aspects of KYC into a single WorkFusion-platformed solution. Capgemini’s UBO process works together on the same platform as Adverse Media screening and Sanctions Screening. Thus, they complement each other. They can integrate with each other to create a single, integrated KYC solution.  

What are the data sources? 
Data Sources are Dunn and Bradstreet, Reuters, and Companies House. Our data model can be customized using internal or external document repositories that you currently use.

At what point will the search be invoked from a KYC and AML perspective? Can this be automated, as usually the AML and KYC checks are automated at the transaction level? Is this solution in place at the time of account opening and maintenance only?
This solution is meant to be used at any point in the process; it is not only meant for account opening. It can be scheduled to run or triggered manually.

What kind of accuracy can we expect from the machine learning engine?
Accuracy is very high for the standardized inputs, such as D&B. Accuracy for a new document repository will start at around 60–70% and build over time as the model is trained.

Does WorkFusion provide API calls so the bank can integrate the results into an existing KYC data collection app?

From a transaction’s Sanctions Screening perspective: What additional value can you add on top of a solution like Fircosoft?
Every AML function within an organization has multiple solutions; Fircosoft is a good example, and there are others tools being used today such as FICA, Finestra and Actimize. Even with these great products though, manual work is required to move between tools and to extract the data to be used for ongoing matching and classification work. Also, Fircosoft, Actmize, etc. rely on structured data, whereas WorkFusion can handle both structured and unstructured sources of data. This will be helpful because the majority of global watchlists have a lot of unstructured data. Our solution offers to take that manual piece of data extraction out, but leave the tools you use in place.

Is this a cloud-based solution or on-premises infrastructure hosted?
It can be hosted on either. We generally see this hosted on-site due to the sensitive nature of the data.

What kind of training data do we need for any implementation?
The models are pre-trained, so you start with a great level of accuracy, and the more you train on your data, the better the model gets.

What version of WorkFusion Intelligent Automation Cloud is required to support this?

Want to learn more about how Capgemini is leveraging WorkFusion’s Intelligent Automation Cloud to reduce AML burdens and create a bigger opportunity for businesses to stay ahead of regulatory requirements? Watch the full conversation and Q&A portion of the webinar, on-demand here.