AI ILLUMINATIONS: How Financial Firms Can Decide What to Automate First


Automation does sound like a good deal to many business leaders in the financial services industry, creating more efficient processes for employees and better experiences for clients—but it also poses a quandary.

How can companies starting on their technology journey identify the right process—or the right parts of processes—to automate? The answer is multifaceted, according to Phil Sheridan, Director of Technology Consultation and Business Development at, an autonomous enterprise platform for financial services organizations.

At, business processes to be automated are described and encompassed in what is called a “process definition document” that outlines every step of the specific process. This is the outcome of involving multiple business stakeholders and subject matter experts. Creating this document involves both virtual and onsite reviews and “getting into the weeds” with those performing the task in question. Without such keen attention to detail, right in the beginning, the project can face many delays down the road.

Then, the process definition document evolves into a “solution design document” that describes how automating a process should benefit the business, offering a path to implementation. At, this document is created by a team of experts across technology and the business.  This ensures a high degree of confidence that the proposed solution will be well received and approved by the client.

According to Sheridan, envisages every automation project in four phases: inception, early phase, growth, and maturity.

“As a firm, you want to start by acknowledging that you are in the inception phase,” he says. “This is where you set up your objectives: What do you want to accomplish with automation? Why are you pursuing this? Which technology partners are you going to engage with to develop a comprehensive automation strategy?”

Moving from inception to the early phase of automation implementation requires some tactical thinking. Sheridan recommends that firms initially target a small number of processes for automation— not more than five initially.  And…. expand from there typically in groups of five.

“The next task when reviewing specific processes at a financial firm—is to map out these processes from start to finish,” he says. “More often than not, if we look at a process in a vacuum, or only a snippet of a process, we actually may overlook the important details.” Within this review, he also mentioned that it is valuable to determine where the information is received to complete the process, and also where the information flows once the process is complete.

Then, before technology is applied, firms and their technology partners should ask if there are ways to simplify the process—are there unnecessary steps involved?  Was there a delightful experience devlivered?

Several firms have adopted long and overly complex processes as “workarounds” for aging technology or other shortcomings, Sheridan says.

We often hear that people are completing a task a certain way because ‘that’s the way it’s been done for the last ten years.’.. Maybe there are legacy systems or compliance reviews, and new regulations have eliminated the need. Further, we also need to consider volume—how many times is this process being performed—to help determine the scale and scope of the project. The end goal of these types of projects is to streamline the process as much as possible and also to make the overall automation as efficient as possible.

The processes that are typically great starting points are the most straightforward ones or those that involve a small number of applications, says Sheridan. “But, we at believe that firms should also consider return-on-investment when selecting where automation should be initially implemented: will automating this process add value across the entire value chain.

If the answer is yes, then there is likely a good opportunity to automate.

At the same time, the ‘human touch’ is equally important for implementing automation successfully.

“It is important for the firm’s leadership to be on board with the project and to be champions within the organization by communicating the overall strategy,” says Sheridan. “If you don’t do that, people may neither understand why they are on these calls about automation nor understand the significance of the project.”  Without this understanding, keeping employees at all levels of the firm engaged in the process can be a tremendous challenge, which could eventually damage the entire project.

The automation experts at also believe that firms need to discuss automation with the people who work the process every day because they know where the weak spots are, where delays happen,  where help is needed, and how improvements can and should be made.

In fact, there are cases wherein a process should be rejected for automation implementation, says Sheridan.

“Does it enhance the overall experience? Does automation add value to an advisor on the field or an internal operations resource? Is what you are achieving through automation improving the overall experience for whomever this impacts? Is automating the process decreasing the risk associated with it? If any of these answers are no, you need to think again carefully before you proceed,” he says. “Automating can make sense from strictly a return-on-investment perspective, but if the return is marginal and if it doesn’t directly improve anyone’s experience, it’s probably not a valuable use of automation resources.”

Firms should also ask whether having a live human associate performing the tasks could add more value to the process. Machines cannot replace human emotions. Having humans engaged with the machines often times will  add greater value.

Firms may also prefer that certain decisions fall under the auspices of human judgment and discretion rather than using an algorithm, Sheridan says.

“If you have too many judgments being made, automation can become overwhelming,” he adds. “It can get complicated quickly, so if you feel there’s value in automating that specific task, maybe you could hold off some of those more complex processes to the later phases as your firm gains more comfort with the automation experience.”

Processes that don’t make sense need to be revised, refined, or eliminated, not automated, says Sheridan.

As firms refine a handful of processes at a time, they should learn from each iteration of automation and become better at identifying opportunities and implementing technology.

“It’s important to realize that in the inception phase you’re going to start slow, and then push for some more,” Sheridan says. “You can start drilling down and scaling out as you work towards becoming a truly autonomous organization.”