When we introduced HealthVerity eXOs, an agentic AI platform powered by Medeloop, in our recent webinar, the discussion quickly focused on how automation can shorten evidence-generation cycles while maintaining transparency and scientific rigor. Our recent webinar featured Andrew Kress, CEO of HealthVerity; Rene Caissie, M.D., founder and CEO of Medeloop; John Ayers, Ph.D., M.A., chief strategy officer at Medeloop; Chris Pashos, Ph.D., head of real-world evidence at Argenx; and Mark Booth head of Portfolio Strategy Rare Disease at Chiesi USA.
Andrew Kress, CEO of HealthVerity, opened the discussion by explaining what has been missing from traditional analytics.
“Before 2016, study analytics were largely manual and non-scalable. Today we are thinking about how to best utilize some of the modern AI-driven tooling to build repeatable pathways that reduce the time from months down to days or even weeks.”
- Andrew Kress, CEO of HealthVerity
He noted that smaller biotech companies often struggle with limited analytic resources. “Many of our customers say, ‘That’s great for the majors, but we’re too small to have a large data-science team,’ or ‘We need data to support evidence generation, but we don’t have the budgets.’ That’s the landscape we’ve seen.” He explained that HealthVerity eXOs was created to address these challenges.
“The goal of this was to use agentic AI to facilitate the operation of studies in a way that would feel very familiar to any researcher that's doing this work manually. But to be able to do so in a way where we can basically allow the researcher to guide the research, and guide the research questions, but allow the tooling to do all the heavy lifting.”
- Andrew Kress, CEO of HealthVerity
Rene Caissie MD, co-founder and CEO of Medeloop, described how HealthVerity eXOs was intentionally designed for interpretability. Each query, regardless of complexity, produces visible cohort definitions, coding logic and outputs. “The mission was clear from the beginning,” he said. “It was to make research as intuitive as asking a natural-language question. We wanted to democratize research to the masses, not just to the experienced researcher or data analyst.” He further noted, “These agents can reason across massive linked data ecosystems. They can design analyses, validate them, and show exactly how they got there. It’s not a black box. It’s full transparency and scientific rigor.”
This design allows analysts and reviewers to understand exactly how results were generated. Dr. Caissie emphasized that transparency is not an afterthought; it is built into the way HealthVerity eXOs operates.
John Ayers expanded on that point, describing how Medeloop builds scientific standards directly into HealthVerity eXOs. He cited measurable benchmarks behind that claim. “When we benchmark against published studies, eXOs recalls about 97 percent of the codes used and identifies roughly 1.3 times more codes that should have been included,” Ayers said. “Our analyses average 93 percent transparency on a 0-to-100 scale, compared to 44 percent in typical studies.”
From the client perspective, Chris Pashos described how Argenx is already seeing the benefits. “Once we start showing efficient answers to questions not in months but in hours, people all of a sudden say, ‘We want more,’ ” he said. “The transparency and reproducibility are delivering. It enables us to move much quicker to more strategic thinking about what we need to understand.” Dr. Pashos added that co-creation with HealthVerity and Medeloop is helping refine both the tool and Argenx’s internal workflows.
Throughout the session, the speakers returned to one message: accelerating evidence generation does not mean lowering scientific standards. By combining verified real-world data from HealthVerity Marketplace with an explainable and traceable AI framework, HealthVerity eXOs allows RWE and HEOR teams to design, prepare, analyze and interpret studies with reproducible logic and defensible results. As the discussion and demo of the tool concluded, attendees were left with an action plan of how agentic AI can support real-world research.
To learn more about how HealthVerity eXOs enables transparent, audit-ready real-world evidence in minutes watch the webinar replay here.