By Aman Kidwai
Workplaces Editor
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
The financial services industry has had to move a little bit more cautiously than some others in its applications of new technology, including automated and agentic software solutions. Because of compliance concerns and entrenched silos in these long-established businesses, new tech opportunities have a high bar to clear to gain adoption. Still, they have been finding success in customer service, internal process automations and skill development.
“We decided to address customer service, and we're doing that from the inside out,” Tim Farrell, SVP, director of business process management at Middlesex Savings Bank in Massachusetts, said at the Druid Symbiosis conference last week. “We have an extensive knowledge base that we're putting in place right now.”
As a 190-year-old bank with a great degree of pride in its community relationships and security, Farrell notes they have never had a data breach, and any new technology proposal meets heavy skepticism from around the leadership team.
Read More on AI
“There was a lot of pushback from some of the higher level executive management team,” Farrell noted, “At the time…there were a lot of different stories out there about how AI was returning inaccurate results, and there was a fair amount of concern.”
He added that AI brings the opportunity for the bank to increase its service standards through proactive identification of potential problems that may arise, for customers or the business.
“Are we identifying positions in the bank that have high turnover? Are we addressing the regulatory concerns that we have on a daily basis? [Such as] money laundering, spending patterns…transactional data that can be analyzed in order to allow us to be able to protect our customers proactively, not reactively,” he told Newsweek in a conversation after his panel.
In the case of customer service, those teams have high turnover that can make it difficult to maintain consistency in the client experience. The agent tool at Middlesex Bank, named Drew, helps employees with onboarding and the "knowledge base" Farrell mentioned which helps them get up to speed on how to do their job and can also interactively answer questions and share relevant information.
“What we wanted to do is we wanted to build a knowledge base that would partner with our people from the inside out, not replace our people,” he said. “It helps give us the opportunity not only to deliver the service and deliver the technology, but to do it in a trustworthy way.”
Farrell and Alwin Bhatia, global head of advisory, capital markets and banking at the IT consulting firm Genpact, both said AI projects designed to reduce costs are often doomed from the start.
“Do not use AI just to reduce your costs,” Bhatia, also on the panel with Farrell, said. “This is a byproduct. If you get a cost reduction, it's great…But AI is too good to just reduce costs.”
He continued: “I think it's really about the business case, the business outcomes you can deliver through that, and they are much more complex than just a simple cost reduction. That’s one of the key conversations we typically have in banks.”
Farrell added that it's important to connect all of his AI projects to the company's bottom line.
“For us, it's always revenue generating, whether I'm training a staff member to be more intelligent by having immediate access to information that's helping to guide their conversation with the customer, hopefully that returns more deposits…we look at it as a revenue stream. We certainly see it as a partnership and a relationship. But again, we're not we're not looking at AI to reduce staff.”
He adds that his team had been working on robotics process automation for about eight years, having saved 56,000 hours.
“We already had built in a lot of efficiencies…So we're very pleased with that, but wanted to take it to the next level,” he said.
The new capabilities of agentic technology have allowed adoption to go smoother and faster in the organization as well.
“It allows us to bring in many facets of data, so we're breaking down barriers and data silos so that we can return more robust and more full answers to questions and queries that our customers are having about their relationship with the bank,” he said. “It helps us from a compliance perspective, from an onboarding perspective, it helps it educate each one of our new team members.
Like many others across industries, Farrell and Bhatia noted the need to manage the change well with employees. Bhatia added that companies should plan for ROI by thinking about what employees can do with the time they save. Farrell shared the anecdote that their AI agent Drew had a diaper on the design, to indicate to employees that they were working with a nascent technology.
“We wanted people to understand it's a technology, but there’s a reason why it's called a language model. It has to learn. It's not just ‘plug it in’ and it just works miraculously, and that's going to be one of the things that we're going to address on a continual basis,” he explained. “Humans aren't always right. So why do we have this assumption that AI is always going to be right? So, we're starting with the knowledge base. We're going to use it to proliferate and break down data silos, and we're going to use it to partner with our staff.”
Having successfully deployed Drew, Farrell sees more agents on the way next year.
“Anytime you deploy technology, it's a challenge on that first project, but once you get through that first project, it's just okay, turn up the next one, follow, repeat, rinse,” he said. “We're going to be very busy going into 2026.”
View 0 Comments
Submit Correction
View Editorial & AI Guidelines