Why your AML system implementation is taking longer than you expected

Discover five leading practices that can be leveraged to mitigate inefficiencies during the automated AML solution process.  

Many organizations encounter numerous hurdles in the process of implementing an automated anti-money laundering (AML) solution for transaction monitoring, customer due diligence and list screening purposes. Deadlines are pushed back, business requirements are changed and issues arise, pushing the actual project delivery date past the predefined timeframe.

In this article, we explore the various reasons for inefficiencies in automated AML solution implementation projects and leading practices for a timely AML solution delivery.

Difficulties in AML implementation

Let’s take a look at a fictional Canadian bank in its journey to implement a consolidated AML monitoring and case management system.

The bank gathered an experienced group of AML professionals, who drafted comprehensive business requirement specifications for the target solution. It then conducted a vendor assessment, and identified and purchased a suitable automated AML monitoring and case management solution.

The bank hired technical resources with relevant experience for the implementation and, after much planning, the project was kicked off. However, the bank experienced numerous difficulties throughout the project, which greatly delayed production rollout, much to the dismay of the stakeholders on the business side.

• Business stakeholders were only given a demo of the system when at the user acceptance testing (UAT) phase. At this point, the business group realized that not all business expectations have been reflected in the business requirements, and as such have not been configured in the system. Certain case management functionalities that the business assumed to be pre-configured in the product were in fact not configured, leading to significant scope changes and additional work. This introduced substantial delays to the project timeline and undue tension between the technical and business groups.
• Relevant data was not identified and sourced soon enough. As a result, technical implementation efforts were held back until the data team could source the relevant data for model validation and testing.
• Technical components were not provisioned ahead of time, leading to significant delays in solution implementation efforts and deployment to higher environments. As a result, testing efforts were delayed, wasting valuable time and resources dedicated to implementation and testing.
• The virtual machines provisioned for the development and testing environments were configured with low processing power and memory in order to manage budgetary constraints. Subsequently, batch processing times were slow, impeding model validation efforts.
• Deployment activities were not automated, leading to significant deployment overhead and challenges in handing over technical implementation procedures to newly on-boarded resources.

The mistakes made in this illustrative scenario can be easily avoided with careful planning and consideration of the various dependencies throughout the project lifecycle.

Consider the following steps to mitigate such errors and facilitate implementation efforts.

Involve the business stakeholders

In many implementation projects, the business users can’t see the system in action until the UAT phase, which may be many months into the project. At this stage, the business stakeholders may decide that a business requirement that was sound in theory doesn’t necessarily work in practice, or perhaps some business requirements that are necessary for the solution to be viable weren’t identified.

This may lead to significant scope changes as well as a substantial reworking of technical components. Such scenarios are better handled earlier in the project, and as such can be mitigated by involving the business stakeholders as soon as the system is implemented and configured with its most basic functions in a development or testing environment.

Frequent demos of case management activities with the target solution should be given so the business stakeholders are aware of all functions exposed by the AML solution and the case management system.

Data first

Identifying the source of each entity described in the AML solutions data requirements document may be a particularly lengthy exercise and should begin as soon as possible. If the data sourcing exercise is started too late, the technical implementers’ work may be blocked, delaying technical implementation efforts.

Provision technical components ahead of time

Provisioning of technical components in a large financial institution can sometimes be a lengthy process. So as to not block technical implementation efforts, the prudent project manager will initiate the provisioning process as soon as reasonably possible.

Development and testing environment components should also be prioritized on the project manager’s agenda. Any delay in acquiring relevant development and testing components might introduce significant delays to the project’s timeline.

Provision the right technical components

For large AML solution implementations involving large sets of data, the testing and model validation processes, as well as pre-production testing, may be slowed down by insufficient resources. Virtual machines with adequate processing power and memory may significantly shorten the duration of such activities.

Adequate configuration should be applied to the container as well as the AML solution so that system resources are properly used when executing batch processes.

Automate the implementation process

A higher degree of automation in the implementation process will translate to greater reusability, improved deployment times and fewer errors. An automated AML deployment solution may be used to track system and solution configuration and automate manual deployment activities.

Conclusion

AML solution implementations can face many challenges. To mitigate those challenges, it is crucial to engage the business stakeholders early and frequently, initiate data analysis and procurement activities ahead of time, and leverage automation in the implementation process. By doing so, you can reduce risk to your project timeline and facilitate a timely delivery.