Q&A with Dr. Jack Prior: Is Your Process Data Ready to Act On?

At the PDA/FDA Joint Regulatory Conference 2026, Jack Prior, head of process monitoring and data, AI strategy at Sanofi, presented the session “Is Your Process Data Ready to Act On? A Balanced Scorecard for Human- and AI-Readiness in Biopharmaceutical Manufacturing.” Following his presentation, BioPharm International® spoke with Prior about what makes process data actionable, where artificial intelligence can add value in biopharmaceutical manufacturing, and what companies need to consider before using AI to support manufacturing decisions.

BioPharm International: In your presentation, you discussed whether process data is ready to act on. What separates data that is simply collected from data that is actually actionable?

Jack Prior: There’s a whole pipeline of things that have to happen from the data you collect to the data you act on. First, you have to collect it and be able to measure it. Then you need to be able to access it.

A lot of our data collection in the industry goes into frontline systems designed for a particular GMP purpose, such as releasing a batch or operating a process. You need to get the data out of those frontline systems, then integrate it and organize it in a way that makes sense. We’re running processes that go through multiple operations and across the globe, between API and fill finish.

You need to trust it. In our industry, everything is highly regulated, so you have to be able to trust it. Then you have to have people focused on analyzing it, and you have to make that easier to do. All those things are barriers on the way to action.

BP: Where do you see AI adding the most real value in process monitoring today, and where is the industry getting ahead of itself?

Prior: I think part of the power is going to be in allowing us to organize our data better and create tools on the fly where we might have built them over years.

Now, something that agentic AI can build for you on the fly is, “Here’s a spreadsheet of some data, here’s a database of some data, here’s a data lake with tables. Help me analyze it.” That kind of pre-work or post-work becomes much easier.

There’s a lot of focus on how we will use it for 24/7, real-time monitoring and control, and we’ll get there. But the first opportunities are going to be in making our engineers more productive, bringing better processes to market, and analyzing the data.

BP: What does a manufacturer need to have in place before AI can reliably support manufacturing decisions?

Prior: You certainly have to have access to data. You need the data from your frontline systems to flow upward to a point where it can be accessed for business purposes and for AI.

Then you need to have, in our regulated environment, some framework by which you’re going to do a risk assessment of how you’re going to use the data. You need to have a good sense of what decisions you’re seeking to make. You can’t just gather it for any particular purpose.

So, the context of use, the risk of any decision you want to make, and then some framework to submit it for approval where necessary.

BP: You also discussed a balanced scorecard for process data in your presentation. What does that approach look like in practice?

Prior: There are six dimensions by which you can look at data. You want it to be fresh, you want to get at it in a frictionless way, you want it accessible to broad groups of people, you want it validated and authentic, you want it structured, and you want it in a standard way.

I think the challenge in the industry today, especially with Big Pharma, is we can put a huge focus on validation. It’s not inappropriate, but it can come at the expense of other things. Or we can put a huge focus on standardization, with that standardization sometimes being done by people less familiar with the actual use of the tools that we need.

So, as process engineers, we need to advocate a little bit more and be a little bit more upfront and clear about what we need, so those needs can be met.

BP: What was your biggest takeaway from this year’s PDA/FDA Joint Regulatory Conference?

Prior: I think there are several representatives here from the FDA speaking about AI with a high degree of fluency. Sometimes there can be a stereotype that the FDA is behind or that they need to be educated. Like all of us, there are opportunities to be educated, but I felt the people here were highly informed, up to speed, and really enthusiastic to see the industry bring forward innovations.

Sometimes we look at innovation and think, “We could do it, but the regulator would never accept it.” I don’t think that’s the case. I think the regulators want innovation, but the rules are also in place. You have to do what we’ve done in the past, do the right things for the patient, and find the new frameworks that are necessary for that.

Click here for more conference coverage.