By Clinical Research News Staff 

July 9, 2026 | Artificial intelligence agents are beginning to reshape clinical trial operations, but sponsors are still hesitating. Raj Indupuri, CEO of eClinical Solutions, joined the latest episode of The Scope of Things, hosted by Deborah Borfitz, to break down the three main obstacles hindering AI adoption: data readiness, governance and trust, and operating model change.  

AI agents cannot compensate for poorly integrated or inconsistent data sources. “If your data foundation is fragmented, an AI agent does not solve the problem,” he notes. As such, modern data infrastructure should be considered a prerequisite for successful deployment. Clinical trials often involve diverse sponsors, study designs, and data modalities, creating complexity that limits AI performance when data foundations are weak. 

For governance and trust, transparency is key. Because AI-generated recommendations can influence patient safety and regulatory submissions, sponsors need clear visibility into how conclusions are reached. According to Indupuri, “checking boxes” is not enough. AI systems need to provide traceable reasoning, data lineage, and audit trails. 

Indupuri described eClinical Solutions' approach as "glass box governance," in which every recommendation can be explained and every human decision recorded within validated workflows. The company introduced its Illuminate AI agents earlier this year across four clinical development areas: data mapping, data review, risk-based quality management, and study operations. While still in the early stages of commercialization, Indupuri said customer interest has been strongest in data mapping and clinical data review. 

Data mapping—converting raw clinical data into standardized formats for submissions and downstream analytics—traditionally requires eight to 12 weeks. With AI-assisted automation, that timeline has already been reduced to two or three weeks, with eClinical Solutions aiming to shorten the process to just days. Similarly, AI-powered data review can dramatically reduce the time required for medical monitors and clinical scientists to analyze increasingly large datasets. By redesigning the operating model, researchers can focus on interpreting findings and making decisions rather than manually searching for patterns. 

Despite these efficiency gains, Indupuri emphasized that AI agents are not designed to replace human expertise. "The goal is not to remove humans from clinical trials. The goal is to stop wasting expertise or expert human work." 

Instead, AI should automate repetitive tasks such as pattern recognition, signal detection, and preliminary data review, while humans retain responsibility for interpretation, judgment, patient safety, and regulatory accountability. 

To learn more about eClinical Solutions’ AI portfolio expansion, as well as the latest news on a two-way participant engagement portal, a KRAS inhibitor trial for metastatic pancreatic cancer, and more, listen to The Scope of Things podcast.