Advarra has introduced Braid, an artificial intelligence (AI) engine designed to improve clinical trial design and operations through access to what the company says is the industry’s largest and most comprehensive clinical trial operations dataset.

The Columbia, Maryland–based research technology and regulatory review company said Braid uses operational data from more than 30,000 studies conducted by 3,500 sponsors, drawn from its institutional review board (IRB) and clinical trial systems. This includes protocol-related documents and operational metrics such as cycle times for startup milestones, enrolment trends, and study modifications.

By curating and integrating this information with standardised ontologies, Advarra says the system can harmonise it with other data sources to deliver “unprecedented analytical and agentic capabilities” for sponsors, contract research organisations (CROs), and clinical sites.

Laura Russell, senior vice president and head of data and AI product development at Advarra, said: “For years, operational data has been buried in static study documents and fragmented eClinical systems, too unstructured to access at scale. Braid changes that. It unlocks operational data across the entire trial lifecycle, capturing important milestones—such as study design modifications—as the study progresses, not just at a single point in time. This allows us to learn lessons from past trials and generate intelligence that improves how research is planned, executed, and scaled.”

According to the company, Braid supports the full trial lifecycle, enabling more coordinated research and reducing the burden on study teams, sites, and patients. It can generate insights to inform trial planning, activate intelligent agents within Advarra’s products and other systems, and automate workflows. Over time, the system is designed to learn from previous applications to increase efficiency and precision.

Gadi Saarony, Advarra’s CEO, said: “Advarra’s unmatched access to trial operations data and deep domain expertise uniquely position us to tackle the complex challenges of creating structured, context-aware data and operationally useful models—models that capture the nuance and precision clinical research demands. While all decisions related to ethical review and participant protection are—and will continue to be—made by people, not AI, we’re actively exploring how AI can enhance Advarra’s systems and operations.”

The first Braid-powered product will focus on study design optimisation, introducing “dynamic protocol engineering” to help align trial plans with real-world operational realities. Advarra said this could include flagging complex visit schedules, identifying eligibility criteria that may slow recruitment, and suggesting realistic startup timelines based on historical site performance.

Saarony added that the company is “building at the speed of trust” and remains committed to developing AI tools that combine innovation with human oversight, scientific rigour, and adherence to ethical and regulatory standards.