What’s New for Regulatory Informatics in the Age of AI? Not Much—And That’s the Point with Raju Rayavarapu of DNAnexus

In this episode of Mendelspod, we’re joined by Raju Rayavarapu, bioinformatician at DNAnexus and former FDA data specialist, to discuss how regulatory science is adapting to the age of AI, big data, and cloud platforms.

* 0:00 How has AI changed regulatory informatics?

* 4:15 KISS

* 12:50 On standards

* 25:00 So nothing new about AI? The juicy question of the interview

* 32:45 Raju’s life: good days and bad nights

Despite the buzz around AI, Raju insists the fundamentals haven’t changed much. “It still essentially follows the same patterns,” he says. “Here's data that needs to be analyzed in order to help make a safety and efficacy decision.” What has changed is scale: more data, more submissions, more complexity. But also more tools.

On that note, Raju makes a strong case for keeping it simple. “It's a really strange choice to do something like a complex language model when all you really need is a string search,” he says. Simpler tools not only reduce costs and computing overhead, they’re more transparent—an essential trait in a regulatory environment where explainability is everything.

Oh, come on, Raju . . . what’s really different about AI? He warns against blind trust in black-box models, especially when trained on biased datasets.

We also explore cloud adoption, standardization, and the global regulatory landscape. Raju explains how DNAnexus is working behind the scenes with agencies around the world to implement trusted environments and validate evolving standards.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.mendelspod.com/subscribe