As CIO of Liberty Mutual, James McGlennon is on a mission to transform the way IT and technology drive business success.
Driving digital transformation for a global insurer that has been in business since 1912 and now employs more than 50,000 people around the world is complex. Luckily, McGlennon has a little help.
“We are fortunate to have 5,000 technologists and an army of software engineers helping us address some of our biggest challenges,” McGlennon said. “Nevertheless, the challenges are diverse.”
James McGlennon
Getting Liberty Mutual into the cloud is no easy task, and developing a digital ecosystem isn’t easy either.
Here, McGlennon, a 2022 MIT Sloan CIO Leadership Award finalist, discusses some of his most pressing technology priorities, talks about Liberty Mutual’s digital transformation and AI explorations, and explains why giving employees flexibility is an important talent management strategy.
Managing complexity as part of digital transformation
What are some of your biggest technical challenges right now?
James McGlennon: The first is technology debt. Many of our solutions are 20 or 30 years old or even older. We’re on a journey to the cloud, and a big part of it is rebuilding new applications in a cloud-native way. In many cases we are looking for ways to run the old workloads on the new platforms. This is a difficult challenge.
The second area we spend a lot of time on these days is what I call data wrangling – how we prepare data for use in our machine learning models and our AI models. We have a lot of data that goes back over 20 or 30 years – sometimes even longer – and not all of it is in the same form.
The third technological challenge is ensuring everything is API-enabled. As we think about how we can participate in global ecosystems with new types of partnerships and partners, it is very important that we can integrate into their world and they can integrate into ours, creating a seamless interaction model.
Including ethical AI
How did you start? to use AI And how has it changed your company?
McGlennon: We started an AI journey maybe four or five years ago, really doubling down on our investment in machine learning and looking at how we could benefit from it. We have also entered into some partnerships with organizations such as MIT, where we conduct research focused on questions such as: Is it possible to create synthetic data to be able to test our hypotheses and models? Can we validate why a model made a particular recommendation? In this new age of AI and machine learning, we need to be able to explain how the models came to their conclusions. And regulators and others will require this as we move forward.
We have a lot of data that goes back over 20 or 30 years – sometimes even longer – and not all of it is in the same form.
James McGlennonCIO, Liberty Mutual
At Liberty Mutual, we have hundreds of machine learning models at various stages of operations. There is hardly a part of our business where we are not deploying some new machine learning models.
For example, we use AI in our operations for our technology infrastructure, which uses machine learning models that can predict when we will have a problem. This allows us to resolve this issue with the platform or solution before it causes an outage or poor experience for our customers.
We have also used AI and machine learning in computer vision technology. If one of our customers has a car accident, they can use their cell phone to take a few photos of the damage to the car and upload them to us. We can feed these images into a machine learning model that we trained on millions of photos of damaged cars. It can help us see the extent of the damage and make an initial repair assessment. In some cases, we can determine immediately whether the car is a total loss so that we can resolve the damage as quickly as possible.
Another point that comes to mind is that we use machine learning engine technology in our claims processing engines. When customers call us, we can intercept them with a bot and perhaps give them a response much faster than if they had to wait for a human.
We’re trying to figure out how we can develop a set of technologies that will allow us to retrain the models when necessary. When you deploy a model to production, the underlying data can sometimes change or differ over time. We need to be able to continuously track the model and retrain it when necessary. This means we are investing heavily in new technologies to automatically deploy machine learning models via API connections that allow us to do this much faster. Speed is therefore important in order to be able to retrain the models across the entire value chain.
Address employees’ need for flexibility
What changes did the pandemic bring to Liberty Mutual and what do you think will happen in the foreseeable future?
McGlennon: We’re focused on how we can ensure that [employees] stay in contact. This was a big challenge for us since we were all virtual. We had to introduce new types of formalized approaches to connecting with people to ensure they felt part of the team and not isolated.
In today’s world, we need to offer our employees more flexibility than ever before. We are leaders in ensuring the safety of our employees and continue to put this at the heart of everything we do.
We have begun a journey to hybrid work. It will probably take a little longer for everyone to get used to coming into the office more often, but our main focus is making sure employees have the flexibility they need. We have been very successful in implementing the things we had to do during the pandemic. We’re not going to bring people back just to tick the “I was in the office” box. It’s really going to focus on the activities that people are doing and the need to come together for certain activities or certain tasks that are personally better.