As Ontario’s largest credit union, and the second largest in Canada, Meridian has a reputation for leading innovation in the country’s financial services industry. When it came time to redesign their processes in consumer lending, they looked at the end-to-end consumer lending journey – from prospect shopping to loan servicing– with a single focus on enhancing the digital experience for their customers.
The Challenge
People have become accustomed to the increasing volume of digital transactions and expect their financial institutions to provide an “Amazon” type of experience – complete with instant credit decisions. Meridian saw this as an opportunity to advance automation and analytical capabilities in tandem with an end-to-end redesign that would make it simpler to adjust risk parameters without complex builds. The challenge in providing a seamless digital and in-person experience was preserving the tenured knowledge of the company’s credit adjudication team members, many of whom have been with the company for more than 25 years.
It was a considerable undertaking, given Meridian’s product suite and project scope, and there were no snap decisions. Internal stakeholders engaged in a rigorous RFP process to identify a risk decisioning platform that would meet Meridian’s needs – both now and into the future. In the end, they selected Provenir, a leader in AI-powered risk decisioning software.
Why Provenir?
“We chose Provenir because of the flexibility and ease of use, combined with the architecture to manage the platform on an ongoing basis and continually add new data sources,” says Van Rooijen. “At present, we’re at that midpoint between legacy processes and future customer expectations. Provenir is allowing us to do that while retaining the knowledge that sets us apart.”
The Meridian-Provenir partnership kicked off in the spring of 2020 to align with the launch of the first product in Meridian’s loan origination system (LOS), a platform designed to handle everything from loan applications and fulfillment to pricing and eligibility. The value of the new decision engines was soon apparent in the ability to make decisions faster and mitigate potential risks, using accurate and real time data.
“The Provenir team understood what we were trying to do,” says Van Rooijen. “Instead of telling us about their solution, they listened to where we wanted to go, dug deep in, and showed us how to get there.”
Advancing the End-to-End Consumer Journey
In the old days, adjudicators might sit in a credit pit surrounded by a group of coworkers and bounce ideas off each other. At the mention of certain deal scenarios, someone in the group would know where to go or who to call to get the information required to make a decision.
Despite the complexity in consumer underwriting, and the complications of COVID, Provenir was able to improve on the insider knowledge approach by providing Meridian with a responsive decisioning tool that configures quickly and adjusts without significant builds and lag times. “That was one of the big things for us and why we went with Provenir,” says Van Rooijen, “We didn’t want to have to build a new cycle every time we needed to change a rule.”
In addition to managing risks more effectively, Meridian now has greater agility to make more informed decisions quickly, improving the member experience throughout the process. The goal, according to Van Rooijen, was to automate as much as possible and then allow Meridian’s adjudicators to focus on the more unique deals – the cases that require a special touch because there is more potential risk.
The Many Advantages of Machine Learning
One of the foundational components of building the organization’s new models was to leverage machine learning to uncover patterns that were second nature to the adjudication team – in some cases patterns that the adjudicators could not identify themselves. This output and optimization of rules brings greater consistency between adjudicators, regardless of whether a customer goes into a branch or interacts digitally. The Provenir platform made it easy to leverage the machine learning output to rewrite a loan or mortgage sample to see what would have happened by tweaking a rule.
“We look at credit differently at Meridian, and our portfolio strength demonstrates we do it really well,” says Van Rooijen. “That’s why it was so important to codify and preserve that special approach to our lending in a way that endures in a scalable end-to-end system.”
