In this video interview, Raviv Pryluk, CEO and co-founder of PhaseV, breaks down the standardization and mapping work that must happen before streaming data can be analyzed meaningfully—and how AI is beginning to make that process fast enough to enable real-time monitoring at scale.
In a recent video interview with Applied Clinical Trials, Raviv Pryluk, CEO and co-founder of PhaseV, discussed the "dead time" problem in clinical data—the gap between when data is measured and when it is clean enough to analyze—and what it will actually take to make real-time trial monitoring a practical reality at scale. He opened by framing the core challenge: the data collection, cleaning, validation, and standardization processes that stand between raw trial data and analysis-ready data create delays that undermine the promise of continuous oversight, unless the right technologies are in place to compress or eliminate that gap.
Pryluk described how data flowing into clinical trials from central labs, CRA queries, EDC systems, and CTMS arrives in formats that require substantial transformation before it meets standards like ADAM and STDM. Until recently, that transformation required weeks of manual work from statistical programmers. The emergence of LLMs and AI-driven tools is changing that, enabling the kind of near-automated standardization that is a prerequisite for real-time monitoring at scale. At the same time, he acknowledged a tension sponsors are actively raising: the risk of surfacing false signals from data that hasn't been properly cleaned. Balancing speed with rigor, he argued, is the central challenge.
On the FDA's
