Aravind Jagannathan, VP, Single Family Data Officer, Freddie Mac, speaks with Chris Knerr, CDO, Syniti about the challenges around data and the role of data products in achieving overall business objectives.

As companies moved to what Jagannathan calls the dev-ops or agile in terms of speed to market in terms of delivery, one of the things that Freddie Mac was looking at was data as a product.

“Our data teams are linked to what you call traditional execution product. So you have an end-to-end holistic delivery model. And that’s also added in terms of our decision modeling of rules capabilities. And why was that important? If it is being viewed as a product, it’s no longer an afterthought as you want to look at your business capabilities and execution model. It’s embedded in everything that you’re trying to prioritize,” Jagannathan says. 

Speaking of challenges around data, he says that sometimes, with governance, it could be viewed as an overhead. Or, in terms of execution, what is the best may not be what needs to be done, while time is of the essence. 

“And then, you start hearing words like tech debt. When you look at our data as a product, we’re moving in sync with the rest of the delivery. And that really has seen the hockey stick in terms of improvements in just a couple of case studies. We re-imagined servicing just as an overall large program. Our capabilities that we’ve delivered this year are going to our end customers,” Jagannathan adds.

To address challenges, he mentions that it is necessary to remove this business technology lens. “It’s really not one team. Folks initially may have been a little hesitant or really embraced it and are living it. So as you’re looking at your release train engineers, it’s not that they’re wearing a business lens or a technology lens, they’ll bring a team lens. And you think about efficiencies, not only from a budget, but speed to market or again business outcomes,” he says.

Sharing his experience of applying AI-ML for business objectives like better anticipating end-customer needs, he says that AI-ML is not an isolated case and is tied to the overall business aspect.

“If I look at those objective key results, it can be tied to, say, one of our features: how am I looking at origination side, shortening our approval cycle? AI-ML is enabled within that. It’s not an isolated case, but it’s tied to that overall business outcome. That’s the key part of data as a product. It is then supporting the overall execution model. So, as I said, AI-ML is not standalone. It’s an enabler to support the overall objective.”

Jagannathan shares that his role intermingles with business as a whole and not just specific areas. And in a scenario as such, it is important to have support.

“I report to our COO who has not only operations, but delivery and execution from a technology division standpoint, and reports to the head of the business. So, I’m in the business area. Having that support from management is also critical for success. For us, having data as a product, being part of that business lifecycle where I’m dealing with my peers, has been successful,” he adds.