Why AI matters at this scale

University of Utah Health Research represents a major academic medical research enterprise embedded within a large health system. With a workforce of 5,001–10,000, it operates at the critical intersection of fundamental biomedical discovery, clinical care, and education. At this scale, the organization generates and manages vast, complex datasets—from genomic sequences and medical images to longitudinal patient records and population health statistics. This data richness is both its greatest asset and a significant challenge. Manual analysis cannot keep pace, creating a bottleneck for innovation. AI and machine learning are no longer optional but essential tools to extract meaningful insights, accelerate the research lifecycle, and translate discoveries into improved patient outcomes more efficiently. For an institution of this size, strategic AI adoption is key to maintaining competitive advantage in securing grants, publishing high-impact science, and attracting premier faculty and students.

Concrete AI Opportunities with ROI Framing

1. Accelerating Translational Research Pipelines

AI can dramatically compress the timeline from bench to bedside. Machine learning models that predict drug-target interactions or simulate clinical trial outcomes can prioritize the most promising laboratory findings for further investment. The ROI is measured in reduced R&D waste, faster initiation of high-value trials, and an increased rate of licensing intellectual property or spawning successful biotech startups, directly contributing to the institution's innovation economy.

2. Optimizing Clinical Research Operations

Recruiting patients for clinical trials is a perennial, costly bottleneck. An AI-driven patient-trial matching system that continuously screens electronic health records can identify eligible participants in near real-time. This directly increases trial enrollment rates, improves study completion timelines, and enhances the institution's appeal as a partner for pharmaceutical companies. The financial return manifests as higher per-patient trial revenue and more successful, on-budget research projects.

3. Enhancing Research Intelligence and Collaboration

An AI research assistant that synthesizes the global scientific literature, internal data, and grant opportunities can serve as a force multiplier for every investigator. By uncovering hidden connections between disparate research fields and suggesting novel collaborations, it increases the novelty and impact of proposals. The ROI is seen in a higher grant application success rate, more interdisciplinary publications, and a stronger overall research portfolio that attracts further funding and talent.

Deployment Risks Specific to This Size Band

Implementing AI at this scale within a complex academic environment carries distinct risks. First, organizational inertia and siloed data are significant hurdles. Research data is often fragmented across departments, labs, and legacy systems, requiring substantial upfront investment in data engineering and governance to create AI-ready datasets. Second, talent acquisition and retention is fiercely competitive. A 5,000–10,000 person organization may struggle to hire and keep enough specialized AI/ML engineers and data scientists, who are often drawn to higher-paying industry roles, risking project delays or oversimplification of models. Third, regulatory and ethical compliance is paramount, especially with patient data. Navigating HIPAA, IRB protocols, and evolving AI-specific regulations adds layers of complexity and potential liability. A failed AI project due to compliance issues can damage institutional reputation and trust. Finally, integrating AI into existing workflows without disrupting core research missions is a change management challenge. Researchers may be skeptical of "black box" models, requiring extensive training and transparent tool design to ensure adoption and sustained use.