Data Management
AI Cybersecurity Europe Fintech Wealthtech Weekend Read

Future Workforce: Emerging AI technologies set to transform the wealth management industry

Wealth management is a data-rich business with a heavy reliance on human data re-keying processes. Whether it is data about clients or data about markets, combining the two intelligently and communicating them to clients is a tenet of the wealth management business model. 

However, thousands of working hours are lost annually at wealth management firms on a multitude of activities. Activities that include poor communication, repetitive tasks, and data interrogation/cleansing are all contributors to inefficiency. The result is a reduction in the business value an individual or team can provide. A significant rise in operational costs, along with inefficiency and compliance issues, are also resulting symptoms. Employees spend too much time on low-value, high-frequency tasks rather than adding value. A value that improves customer experience or provides new revenue streams.

Stuart Robson, Financial Services Business Analyst, Future Workforce
Stuart Robson, Financial Services Business Analyst, Future Workforce

Technology has always been viewed as a way to combat this. But only recently have advances in artificial intelligence (AI) been made that drastically change the dynamics. This article, provided by Stuart Robson, a Financial Services Business Analyst at the robotic process automation (RPA) company Future Workforce, investigates some emerging technologies and how they will transform an organisation’s capabilities to improve scalability and improve customer experience in a more beneficial timeline than performing numerous traditional IT projects:

Let us first understand a little more deeply some of the issues faced by Asset Management firms today. Asset management firms tend to have a segregated structure (notwithstanding those areas segregated due to financial regulations). Different teams handle different products or tasks. And with finance being such a broad and deep field, people tend to build their experience or specialise in particular subfields, similar to medicine in some respects. 

This segregation is also true of the applications employees use, and these develop iteratively over time to meet very niche needs. For example, one application is used by the performance team to calculate portfolio returns. Still, a different application is used by the client reporting team to produce reports that contain those returns. 

There are advantages and disadvantages to this. First, each application is tasked with meeting only a limited set of criteria and developed by subject matter experts in that area, thus fulfilling those criteria very well. However, rarely, if ever, are they a plug-and-play solution. Systems often cannot interact with each other and use incompatible data formats. Thus, employees must become the go-between, manually duplicating data between systems, i.e. a performance analyst emailing portfolio returns to a member of the client reporting team.

This process is lengthy and presents many operational risks, including keying-in errors. Further manual work is then required to investigate and correct mistakes. Client deadlines are affected, and SLAs (service-level agreements) breached. All of which can lead to immediate monetary loss, with penalties written into those agreements and future revenue loss due to reputational damage.  

So, what innovations have been happening in the world of artificial intelligence to improve business outcomes? The highest gains are in the intelligent automation field. Technologies such as process mining, robotic process automation (RPA), conversational AI, and automated decision-making deliver real outcome improvements to financial institutions.

RPA software can be utilised to transform legacy processes without replacing or redeveloping legacy systems that lack APIs or databases. For example, UiPath has zero-code machine learning capability, which means a person without coding skills can build and teach a software robot to do simple tasks and assume the role of data scientist to automate more complex processes.

Such a suite of technologies is combined to complete manual tasks faster and more consistently than a person and represents the greatest improvement in time to value. Imagine having a robot that could be utilised to read performance returns from one application and enter them into the client reporting application without error. Aside from the obvious benefits this provides, it also removes a mundane task from an employee’s day and improves satisfaction, allowing them to focus on more strategic work.

Another excellent machine learning tool is Re:infer, which performs communications mining. Its deep learning technology converts unstructured communications (emails, calls, chats, notes) into structured data in real-time. It does this by unleashing its unsupervised learning algorithms on a source of unstructured data, e.g. an email inbox. These then create clusters of similar communications (emails in this example) to be presented back to the ‘model trainer’ or user.

The software cleverly groups emails it thinks are about the same thing, all by itself. It then asks the user to label the cluster in their business language to learn what conversation these emails are covering. The tool will then present back more examples to be confirmed and continuously reinforce its learning. The user can also create labels for emails within a cluster, giving the tool a more granular understanding of their content, ensuring any single email does not have an unhelpful label.

With machine learning tools, the user can also label data points within emails by simply highlighting an entity and giving it a name in seconds. And this is where automation possibilities are unlocked. You can educate the machine learning tools to identify what a Sedol, ISIN, gross return, or net return looks like from within the body of an email. Company name, client name, fund name, and more. The tool will then find each entity’s nuances to present back to the user for affirmation/correction. Unlike other machine learning software Re:infer does not require thousands of data points to begin learning; less than ten will suffice. Which means you start seeing benefits straight away.

One obvious area of a wealth management company where this could make immediate efficiencies is handling IT requests. Using metadata gathered from communications mining, the number of incoming requests relating to a particular area, and the number of emails it takes to resolve, the software helps identify where automation opportunities should be explored first. Or where the most significant automation benefits could be realised.

A high volume of IT requests relate to onboarding in many organisations, whether a new client or a new employee, to access various applications or databases. A machine learning product will understand an onboarding request coming into the IT mailbox. It would also be able to determine if it is a new client or a new starter. This is where a product’s ability to recognise and pick out data elements comes into play. Coupled with RPA software or even simply communicating directly API to API, the application can open a workflow tool, such as ServiceNow, and open a ticket for that onboarding request, completing the necessary fields with the information contained within that email.

Using RPA capabilities, robots then communicate to relevant teams and systems, setting up the client or worker correctly. The whole end-to-end process can be automated using a suitable workflow tool that manages how each process is handled and by whom. Based on rules and actions specified by the user, i.e. which robot sets the client up in a particular system or sets up access for a new user, or which teams are sent notifications or calls to action, further freeing up time to concentrate on higher-value work.

The future of wealth management is a highly competitive landscape with new entrants leveraging the latest technologies to gain business model advantage: attracting new customers and achieving business growth while keeping costs low. To compete, modern wealth management firms need to invest in the right digital capabilities. Firms that successfully blend AI with human expertise will see the most significant results. When applied individually, these technologies drive major improvements in efficiency, service, and business agility. In unison, the possibilities are endless.


  • Tyler is a fintech journalist with specific interests in online banking and emerging AI technologies. He began his career writing with a plethora of national and international publications.

Related posts

Extension on Strong Customer Authentication Changes Welcomed by UK Banks

Tyler Pathe

Open Finance Association Launched to Drive Open Finance Across Europe

Francis Bignell

UK Fintech News Roundup: The Latest Stories 09/11

Tom Bleach