Insights from Comerica & Infosys

Chad Watt: Welcome to Ahead in the Cloud where technology and leadership intersect to shape the future of industries. I'm Chad Watt, Infosys Knowledge Institute, researcher and writer. Today I'm joined by Yar Savchenko, Senior Vice President of Infrastructure and Cloud Operations at Comerica Bank and Shripad Shanbhag, Infosys Vice President and Delivery Head for Cloud Infrastructure and Security Services. Yar brings deep experience in financial technology from previous leadership roles at Capital One and State Farm. Since joining Comerica in early 2024, he's been focused on strengthening the bank's technology foundations while preparing for what's next in banking infrastructure. Shripad brings more than three decades of technology management and delivery experience at Infosys. In his current role. He has the delivery organization for clients in the financial services, insurance, healthcare, and life sciences segments in America. He is based out of the company's headquarters in Bengaluru, India. Gentlemen, thanks for joining us.

Yar Savchenko: Thank you. Excited to be here.

Chad Watt: I would say in 2024, we had a number of challenges that we worked on and successfully addressed. Some of them include migrating some of the legacy applications into the cloud and making them cloud-native. In addition to that, we also had several other efforts associated with strengthening our infrastructure to ensure it's more resilient, utilizing cost-effective solutions. And I'm sure we'll touch on this a little bit later, but cloud is very expensive if not done properly. And then the third really big piece for us, being a bank, we are tasked with protecting our customer's money. We want to ensure that our infrastructure is as secure as possible and that we are taking good care of our customer's money.

Chad Watt: Great. Let me follow up on that, Yar. What is distinct about IT and financial services?

Yar Savchenko: That's a really good question. I would say that there's a very clear line in the sand when it comes to two. IT is focused on bringing new features as fast as possible to their customers. In the banking sector, you want to make sure that you are bringing new features, however, at a pace where you're not introducing any potential to [inaudible 00:02:20] issues and customers are able to get to their money when they need to the most.

Chad Watt: Shripad, can you add to that?

Shripad Shanbhag: Financial services industry has been at the forefront of adopting technology to make their infrastructure secure, scalable, and available. Cloud has been a great enabler in ensuring that the technology becomes available to customers, and it also is scalable. Compliance and security has been a great inhibitor in adopting cloud for our customers. However, hyperscalers over the year have brought in technology improvements so that it becomes secure and it is something which they can rely on for taking their workloads on cloud.

Chad Watt: Yar, how do you add capabilities via cloud and infrastructure while also keeping that cost in check?

Yar Savchenko: There's a couple of ways that a company, especially mid-sized banks, such as Comerica, can effectively utilize cloud technologies. One perspective is to do what's known as lift-and-shift, where you take your existing application from where it resides within your on-prem data center and you move it into the cloud without doing any redesign or re-architecture. That is usually what leads to having a very expensive bill for your cloud infrastructure as well as some stability or reliability issues. The second and more mature way of doing it is when you take an existing application and you re-architecture it to utilize a microservices architecture, where you take a monolith application and you break that into separate microservices.

And that will not only allow you to right-size them, and you'll hear me use the right-size a lot, but that is ensuring that you're appropriately sizing on your microservices within the cloud, but also use cloud elasticity features. And that is one of the main selling points of the cloud is if you look at some of the cloud providers such as AWS, Azure, or GCP, they have almost infinite amount of resources and you can scale up your microservices as much as needed or as little as needed to save cost in the long run.

Chad Watt: Shripad at a high level, can you talk to me about FinOps and how you put that into practice?

Shripad Shanbhag: FinOps is all about managing the costs of utilizing the cloud resources. And everybody in the organization has a role to play in optimizing the costs. It is not only the people who are providing the infrastructure, but it's also everybody in the organization, including the architects who are architecting the systems so that they design and framework, and architecture required to manage this properly. Secondly, the operations, the teams which are managing the operations as well have a huge role to play in terms of monitoring and metering consumption of cloud. And lastly, it's also important to ensure that there is no wastage of resources, and the cost is allocated to the right department, which is consuming it. So FinOps practices is not a central team's responsibility, it is something which everybody in the organization has to play a role into.

Chad Watt: Looking ahead to the rest of 2025 and beyond, what's top of mind for your infrastructure roadmap?

Yar Savchenko: I would say for us in Comerica, it's making sure that we continue migrating some of remaining applications into the cloud. We utilize as many of those cloud native features as possible for our applications, again, to make them elastic resilient as well as ensure that we are applying appropriate financial management to our cloud resources. And lastly, I would say for us, it's also ensuring that we pursue with developing and delivering features that our customers are looking for. And those can be real-time access, those could be ability to access your account anytime from mobile or web applications. And as the IT industry overall matures, those ways of interacting with your financial institutions become more and more critical every day.

Chad Watt: Shripad, what do you think of those goals?

Shripad Shanbhag: I think these goals are very important for Comerica, and I believe they're realistic and they're achievable. These goals in terms of consolidation, automation, and standardization are going to be extremely important. As part of Infosys Cobalt repository, we have a huge set of reusable assets. That's something which we can bring as an added toolset that Comerica can utilize. One of the recommendations we have is also that how AI can be utilized into operations to bring down the human intervention and the effort human beings put in solving mundane tasks. That's something which we believe that can be brought in to improve further in the next year. But I'm really looking forward to enabling your strategic roadmap and your priorities in this year.

Chad Watt: You mentioned some pretty ambitious goals, migration, automation, vulnerable new management, new features for clients. What does the future state of infrastructure at Comerica look like in your mind?

Yar Savchenko: I would say, and it sounds like I'm going to be repeating myself a little bit, but taking advantage of the cloud features and utilizing cloud-native applications that were rebuilt from the ground up to fully utilize the cloud, I would say, pretty much unlimited capability that's available to us. The other factor that I would mention is as you are migrating to the cloud, the experience and expertise of your engineers is extremely important.

Migrating into the cloud requires a brand new skill set that sometimes might be available in the market, sometimes it might be achieved by reskilling or upskilling your existing engineers. And I don't think we mentioned this in our previous answers, but I think that's another key aspect where technology continues to advance at an amazing rate, right? If you look at AI, as you mentioned, right? That is a brand new, a whole large aspect of IT industry. And making sure that the engineers that build support and manage all of the infrastructure components are up-to-date and have the latest and greatest skill set is what can help any company to stay ahead of their market and competition, and to ensure that their environment is as resilient as it can possibly be.

Chad Watt: Moving from lift and shift to cloud native development, moving from development to automated AI driven development, how do you go about delivering that?

Shripad Shanbhag: So two things, one is to see how we can bring the best of tools and technologies made available to clients like Comerica. And secondly, how we can make the people ready. Because like Yar was saying, people readiness to support these technologies is as important as the technology itself, right? How can we make sure the people are available and the tools and the thought process is brought in to implement? And lastly, like I mentioned earlier, Infosys supports huge set of clients across the globe. We can bring in the best practices, tools, reusable assets, to further this transformation. That is how we can support Comerica and the clients, and that's what we have been doing all these years.

Chad Watt: Yar, we've talked about the value of resilience. What's the best way to ensure resilience in your IT organization?

Yar Savchenko: I would say that the best way to stress test your system and ensure that the resiliency operates the way you design this is chaos engineering. It's something that Netflix introduced over 10 years ago, and what chaos engineering is all about is to introduce failure in production during high volume date where you have the most customer transactions, and then stress test your systems in a controlled environment to ensure that resiliency features you have in place, whether it's in a cloud or on-prem, are truly working as designed. It is a lot better to identify those gaps within your resiliency, protections, and defense in a controlled environment where if something fails, you can bring it back up and restore services as best as possible, versus finding out that a resiliency feature doesn't work when you're in middle of an outage.

Chad Watt: Shripad, if this works so well, why isn't everyone doing it?

Shripad Shanbhag: It's a very good question, Chad. In theory, it looks very good, but if you have to implement it, it takes a lot of effort and it's really scary, right? And also people don't want to know what is wrong with their systems, yeah, that's another part of it. My recommendation is if you have to really do it, pick up the critical business flows and try with that, and get success with it and take it to the broader system so that you can progressively move towards the chaos engineering practices, which Yar spoke about. But in my view, it is a fantastic thing to do. A lot of organization could have taken benefit for whatever reason, they're not [inaudible 00:12:11], but there is definitely merit in implementing it.

Chad Watt: So gentlemen, thank you for your time today. Can I ask you for your top three takeaways from this discussion?

Yar Savchenko: Definitely. Thank you so much, Chad. So I will start before I pass this on to you, but for me, if I had to summarize our conversation down to three main points, one would be making sure that your workforce is future and cloud ready. Developing the skills and make sure that those skills are applied on a daily basis. Two, is migrating your application into the cloud and redesigning and re-architecturing them to take advantage of all those cloud features that are being offered to you. And three, I would say, this is the most important regardless of what industry you're in, whether it's IT or financial, is to ensure that your infrastructure, the applications that [inaudible 00:13:01] that infrastructure is resilient as possible.

Shripad Shanbhag: Yeah. So thanks Chad and Yar for this wonderful conversation. I just want to say that two, three things I want to add before concluding. One is, at the end of the day, we are here to service our end customers, right? We can bring the best of people and technology, but what matters at the end is what experience we provide to our customers. So, that's what we are going towards in terms of providing better experience to good customers. Second, how can we actually make the system resilient and recover fast so that we can reduce the impact to business? So those are the couple of points in addition to what Yar talked about. But otherwise, we are at an exciting time where technology is changing, client expectations are changing. So I think the priorities, which you laid out, we are really looking forward to working with you in this year, and I really want to thank you, Yar, for your time.

Yar Savchenko: Definitely. Thank you so much for your partnership. And Chad, thank you so much for hosting.

Chad Watt: Great. Thank you guys very much. This podcast is presented by Infosys in partnership with MIT Tech Review. Visit our enterprise AI Hub at technologyreview.com to learn more, and be sure to follow wherever you get your podcasts. You can find details in our show notes and transcripts at infosys.com/iki in our podcast section. Thanks to our producers, Christine Calhoun and Yulia De Bari. Dode Bigley is our audio technician. I'm Chad Watt at the Infosys Knowledge Institute signing off. Until next time, keep learning and keep sharing.