When New York Life was founded in 1845, the cutting-edge tool of the day was the rubber band. Information technology has evolved a tad since then, and Bill Cassidy's job is to bring the insurance company into the 21st century.
"We're the proud stewards of a 175-year-plus company with a deep history," said Cassidy, chief information officer, in a recent interview. "But you know, we're not dissimilar from large organizations with a fair amount of technical debt that we're trying to remediate, and, at the same time, we're trying to create modern experiences for our agents and our customers."
New York Life is in the middle of what Cassidy called "a multiyear strategy" to modernize its technology infrastructure and applications. The plan is focused on three key areas: cloud computing, cybersecurity, and data management.
Like many companies over the last decade, New York Life is trying to get out of the self-managed data-center era and move its corporate applications to newer infrastructure.
When the project is complete, "we envision about a third of our applications will be in our (colocation) facility, a third of our applications will be in our cloud environment, and a third of our applications will be hosted (by) the SaaS provider," Cassidy said. New York Life uses private tenants on AWS for its cloud computing needs and selects colocation facilities around the country to fill in the gaps between AWS regions as needed.
Moving software designed for the mainframe is a tricky process. The company is still running "multiple critical applications" on mainframes, Cassidy said, and while New York Life eventually plans to replace those apps with more modern versions, he still thinks mainframes have a role to play in modern enterprise computing.
"Our policy administration application is homegrown. It's not a highly performing application because it runs on the mainframe, it's not a highly performing application because it was built 60 years ago for a different set of use cases," he said. Very few companies are creating new applications for mainframes, of course, which is why New York Life's decision to wind down its use of mainframes is more about being able to provide the types of modern applications its customers and agents want.
"Last time I checked at MIT they're not teaching COBOL, more than likely, as a primary language," Cassidy said. But there are more people out there with COBOL skills than you might think, he said, and by augmenting that talent with help from consultants and third-party contractors, New York Life has been able to plot out a timeline for switching from COBOL apps to modern custom-built or SaaS apps.
Cassidy's next concern was modernizing New York Life's approach to cybersecurity.
"If you are not thinking about modernizing your cybersecurity capabilities in concert with moving your compute and data into the cloud, you'll find yourself digging yourself a hole that you're not going to be very happy in," he said.
There was a simplicity to security in the self-managed data center era, to some extent: Stop the bad guys at the perimeter of the network. That doesn't really work in the cloud, where the emphasis has shifted to identity management and governance around which employees are trusted with access to sensitive corporate data.
This is obviously a key concern at any financial institution, and New York Life operates its own security operations center to detect and monitor its systems for any anomalies using tools from Zscaler and IBM. The company partnered with Deloitte to staff that security center and conduct the actual monitoring around the clock, which allowed its in-house security professionals to focus on more strategic problems, Cassidy said.
"This is one of those good examples that if you pick the right strategy, technologically, it also opens up an opportunity to use your talent in a much more creative way," he said.
Mining the insights
Insurance has always been a data-driven business.
"We collect, consume, and store unbelievably large volumes of data," Cassidy said. However, "data quality and insights from data is not something that you achieve by chasing data itself, you get the data and validate the insights by fixing other fundamental things," first and foremost making sure that business goals and technology goals are developed in concert, he said.
New York Life relies on several AWS tools for data management and uses MuleSoft to connect its data with its applications. Data from other Salesforce products — including the flagship CRM tool as well as Tableau — is presented to company executives on a near-constant basis to help inform decisions.
The real challenge with data at the moment, however, isn't so much about managing and storing the data but making sure you get something out of it, Cassidy said.
"I think really what the step function change is going to be is on the insight generation, not on the data management side. And this is where the AI comes in," he said.
New York Life has plunged into the generative AI era, according to Cassidy. The company has set a few internal restrictions around how employees use tools like ChatGPT — "there's a cost component," he wryly noted — but for the most part is encouraging its workforce to use generative AI tools to solve business problems and move more quickly.
"We have a separate and distinct team that's obviously driven by our business partners to identify the most beneficial use cases that we could apply AI to," he said. "We are fundamentally challenging the multiyear strategy we have from a business perspective to see if there's a different way that we can look at solving those problems through generative AI in the interest of leapfrogging and accelerating our modernization."
Correction: This story was updated to correct the prevalence of Salesforce applications within New York Life's tech stack.