In what is a recurring feature, Finopotamus will profile interesting and intriguing tech professionals who are positively impacting the credit union industry.
For this issue, we visited with Modelshop’s CEO Tom Tobin. Designed for risk modelers by risk modelers with more than 30 years of industry experience, Modelshop’s no-code, artificial intelligence (AI) risk decision platform has been used by clients to design, test, deploy and monitor new risk models without having to “rip and replace” existing tech stacks. Today, Modelshop powers the risk and decision-making software for some of the world’s most innovative financial institutions, credit unions, banks, and private lenders.
By W.B. King
Back in 1979, 13-yeard-old Tom Tobin’s life forever changed when he was introduced to a Sinclair computer kit recently built by his uncle. “I was immediately hooked on the power of creating something from just an idea and a keyboard,” Tobin told Finopotamus. “Since then I’ve been focused on re-creating that excitement over and over.”
After earning an Associate’s Degree in Engineering from Dutchess Community College, Tobin next received a Bachelor of Science in Electrical Engineering from Cornell University. His resumé includes serving as a senior developer and architect at General Dynamics, where he helped develop a simulation of a machine vision Space Station docking system for the Atlas-Centaur upper stage rocket. He next worked as a vice president of development at HNC Software, which was acquired by FICO in 2002. Here he championed the funding, development and roll-out of a new, modern predictive modeling platform for internal use and external sale.
“I’ve studied electrical engineering to understand how computers worked, built avionics for space vehicles, created machine vision technology and worked on the precursors to artificial intelligence (AI) back in the 90’s,” he shared. “Each adventure held that alluring promise that with some creativity and technology you can create anything you imagine.”
The Wild West of Early Startups
Prior to founding the Newark, N.J.-based Modelshop in 2014, Tobin worked at Oracle as a senior vice president of products, the CTO of Fidex, and at Fiserv in two capacities: CTO of development, risk and compliance and general manager of its financial crime risk management solutions.
“Over the course of my career the biggest trends I’ve seen are a shift towards a more mature and responsible technology culture compared to the wild west of early startups. I’ve also seen much more of a reliance on technology re-use, whether that’s components and frameworks or entire platforms that can accelerate the development of new solutions,” he said. “The days of building everything from scratch are rapidly ending. I’m happy to see that culture expand into financial institutions, including credit unions.”
With 11 employees, seven of which are tech-facing, Modelshop’s ethos is based around “capturing the thrill of innovation,” Tobin told Finopotamus.
“I look for that culture when I’m hiring as well. If I see a resume that only describes the nuts and bolts of the technical work someone did, it’s a pass for me,” he shared. “I want to see why a technology candidate was excited by what they helped create.”
Tobin also subscribes to another management philosophy: encouraging his technology teams to think like a business user.
“It’s easy to create terrible software if the developers don’t take joy in using what they’ve created,” he continued. “For that reason, every developer is also a product manager and a quality assurance person. They need to own their creation and work with the business to make sure the software they write solves the business problem.”
A Singular Focus on the Member Experience
In Tobin’s view, what separates credit unions from other financial institutions (FIs) is the collaborative nature of its technology teams. This informed comparison comes as a result of working for many years with large FIs, where “technology teams acted as gate-keepers, often not trusting vendors,” he noted.
“The open and collaborative IT teams I’ve met at credit unions have been refreshing. I believe this culture starts at the top with a shared vision of serving the members,” said Tobin, who added that Modelshop recently signed one of the top 10 credit unions in the industry as a client.
“As a result, IT teams are less interested in building everything from scratch and prefer to leverage the best technology available to deliver member value faster,” he continued. “I’ve also been impressed by how well the business and IT teams work together with a singular focus on making the lives of their members, and their employees, better. It seems like that’s the way organizations should work, but we all know that’s not always the case.”
When it comes to pain points related to a credit union’s credit-decisioning platform, Tobin said that a leading issue is the disconnect in how the risk team and production IT team work with and analyze data points. Respective objectives, tools and processes between these teams, he added, are different.
“Risk teams want to touch and feel the data, understand the relationships and correlations and quickly introduce new data and variables that can help them make better decisions,” Tobin said. “Because of this, the risk teams use tools that encourage that flexibility, spreadsheets, BI (business intelligence) tools and analytic languages, such as Python and SAS.”
Conversely, he offered that “production IT teams are completely focused on stability, security and control.” To this end, these team members view data as “transactions flowing through the process.” As such, they want to “lock down variation” in how transactions are processed.
Aligning Data Cultures
Introducing new data sources or variables into the production system is often a significant engineering effort, which could take months to execute, Tobin noted. This is a result of these “data cultures being diametrically opposed to each other,” which causes friction, he said, adding that each of these teams, if open-minded, could draw inspiration and innovation from the other.
“Risk teams’ lives would be much easier if they could leverage the structure and automation provided by IT managed systems, and the IT teams would be a lot more agile if they had more flexibility in the data they could introduce and the variables they could create in the production system,” he continued. “If these teams could use the same tools to work with data and variables they could skip the time consuming and risky re-coding of logic as credit decision engines move out of design and into production.”
Over the last few years, Tobin has seen more mutually beneficial partnerships between credit unions and fintechs, a trend he strongly supports. There are multiple reasons for this evolution of shared thoughts and goals, he offered.
“For one, fintechs have begun to move away from being challengers to established FIs and are increasingly applying their innovative technology to help traditional financial institutions. Another reason is that the pace of innovation has been accelerating and credit unions realize that they need help keeping up,” Tobin continued. “Fintechs have the agility to develop new technologies quickly and to mature them by working with multiple credit unions. The spirit of cooperation among credit unions makes cross-credit union learning facilitated by fintechs even more powerful.”