The Bank of Montreal is on a digital journey that it measures beyond the traditional cost-related key performance indicators (KPIs) and uses two cloud providers to do so.

The $805.9 billion bank focuses on measurements like talent rescaling and automation capacity, Victor Tung, U.S. chief technology and operations officer and chief information and operations officer, BMO Capital Markets, said during a panel discussion at the Bank Automation Summit.

“An interesting spin on how we actually do automation is a measure of how many people actually scale really quick with this new skill set and how do we actually turn over the talent base,” Tung said. “We look at how many people we actually hire into right skill sets. We look at our attrition rates … and we look at how do we actually build a foundation for the future in terms of the skill sets that we actually need.”

The Montreal-based BMO uses automation on its target platforms and monitors the metrics in the tools’ stack. But evaluating how much automation the bank derives from homegrown activities is key, Tung said.

While some banks are using one cloud solution, BMO is taking a multi-cloud approach, deploying on both Amazon Web Services (AWS) and Microsoft Azure, he said.

Containerization as an automation tool

While Tung did not elaborate on Azure, for AWS he said the bank uses Lamba, an event-driven, serverless computing platform. Lamba allows organizations to run code without provisioning or managing infrastructure, according to AWS.

Lamba supports containerization, and understanding how it works is key to determining how the cloud can help IT automate.

Containerization is used for cloud-native applications and works by “packaging of software code with just the operating system (OS) libraries and dependencies required to run the code to create a single lightweight executable — called a ‘container’ — that runs consistently on any infrastructure,” according to IBM. It’s like a virtual machine but more portable and resource-efficient, IBM added.

Essentially containerization allows IT to try out code without having to build a special infrastructure for testing and deployment at scale.

Often cloud conversations will refer to Kubernetes, which is a Google-created, open-source container orchestration system. But containerization itself is an approach that has developed over decades of computer science, David Snitkof, vice president of analytics at Ocrolus, explained to attendees at the Summit.

“The way to think about containerization is you’re giving the developer all the tools they needed, internalizing the dependencies, not having the other dependencies so that developers who work for you or for me can build something and deploy to production with a minimal amount of support from other people needed,” Snitkof said.

Even so, it’s tricky, Tung said, when you’re a 200-year-old bank.

“If I was in a new company, I could build everything from scratch,” Tung said. It becomes more challenging when a bank tries to bridge legacy systems with new technology and plug it all into the same ecosystem, he explained.

“That comes back to my first point of the measure around talented reskilling; that’s why it’s so important,” Tung said.