Founded in 1784 by Alexander Hamilton, BNY Mellon is one of the oldest banks in the U.S. and is the world’s largest custodian bank and securities services company, with $2.4 trillion in assets under management, another $46 trillion in assets under custody, and more than $307 billion in private wealth.
It is also evolving to become a digital bank, with cloud a key element of this transformation.
Laying the foundation for its cloud strategy, BNY Mellon undertook a multi-year application modernization effort. “In that course of that journey, we virtualized and containerized approximately 95% of our distributed applications in our internal ecosystem. We essentially built an orchestration layer and viewed public cloud as just another landing zone outside our data centers,” says Joe Sieczkowski, CIO of architecture and engineering at BNY Mellon.
At CIO’s recent Future of Cloud Summit, John Gallant, enterprise consulting director with Foundry sat down with Sieczkowski to learn more about his cloud strategy, governance in the cloud, and leveraging cloud where it is most effective. What follows are edited excerpts of that conversation. For more insights, watch the full video embedded below.
On BNY Mellon’s cloud strategy:
First and foremost, we view cloud as a journey, not a destination. Our strategy is essentially to leverage the public cloud’s economies of scale, to drive business value, to lower risk, to increase resiliency, and really to ensure our infrastructure is evergreen. Essentially, cloud enables us to better serve our stakeholders.
Today, our strategy is to have a multi-cloud approach. We have to go to where our clients are. We are going to pick best of breed solutions and maintain our ability to pivot as needed. We are going to limit lock-in, we are going to understand our exit strategy. And frankly, for critical business workloads, we actually may select a process on multiple providers, like Azure and GCP or AWS and Azure, etc.
Governance in multicloud environments:
BNY Mellon already has a rigorous governance process. And our approach has been to extend that process and enhance that process to cover cloud. So as an example, every development initiative has to go through a permit-to-design, permit-to-build, and permit-to-operate tollgate. And that is where we do the architecture reviews, the security reviews, the risk reviews, and even the operational reviews to make sure that we are appropriately securing, monitoring, and governing everything we do for our stakeholders.
BNY Mellon’s modernization journey:
BNY Mellon is evolving into a digital bank. The key point here is that our cloud strategy is part of our overall technology and digital journey, as we continually modernize. So we view this as we have laid the foundation for public cloud with our internal modernization journey. This included enhancing our designs, our standards, our controls, quality assurance, as well as the governance and tollgates around it. We have established well-defined designed patterns and blueprints that are constantly evolving, and we also established anti-patterns that people must avoid. Technology is always evolving, and we have to evolve with it and continue to manage it professionally.
How cloud bolsters resilience:
BNY Mellon has a very strong resiliency posture. However, we believe cloud will afford us an opportunity to really think about next-generation resiliency. This includes scaling during market events, avoiding missing windows, avoiding missing service-level agreements. And frankly, we have also been thinking about the notion of a lifeboat in the cloud, meaning that if there was a really drastic event, we could spin up a lifeboat—in the cloud—to process workloads. We are thinking about it as a cost-effective way to further improve our resiliency posture.
Where cloud is most effective:
One area I think cloud is just going to be really effective is any area which involves experimentation and has a high opportunity cost. Because when you can experiment, you can potentially enter a new business quickly, test an idea. So, for instance, let’s say I have an idea or one of our leading data scientists has an idea for a next-generation fraud model. We can spin up 1,000 GPUs in the cloud for 2 weeks, test a new fraud model, and then turn them off. I just completely removed the opportunity cost.
Another thing that comes to mind is there are customers that want the data close to them. And so cloud in another area, if a customer happens to want their data in their own country—both for latency reasons and perhaps for data domicile reasons—it allows us, as an organization, to go to the customer rather than have the customer come to us.
On AI in the enterprise:
I believe data science, ML, and AI will actually transform the enterprise. In and outside of financial services. In my point of view, regardless of industry, effective firms will be deriving insights from data and driving actionable strategies to provide value for the customers and stakeholders.
At the end of the day, AI and ML isn’t magic. At its core, it is sophisticated and complex math on data. And you need to understand purpose and outcome, establish the right due diligence, peer review, and your processes to test your algos, to ensure effectiveness.
And you hear this a lot. But at the end of the day, you need to make sure your results are explainable and free of bias. It is all about data driving insights and insights driving strategy.
