Case Study
April 2026
How Duke Clinical Research Institute partnered with Health Universe to condense a 6-9 month process into 7.5 days, without sacrificing scientific rigor or regulatory integrity.
Health Universe's AI infrastructure and multi-agent orchestration system reduced months of operational complexity into days, while maintaining robust audit trails and compliance standards required by academic medicine.
30-40x
Faster clinical trial setup
93%
Time savings across key workflow tasks
10x
Faster document processing
Duke Clinical Research Institute
DCRI, part of the Duke University School of Medicine, combines the clinical expertise and academic leadership of a premier teaching hospital with the capabilities of a full-service clinical research organization.
i-Cubed
Imagine, Incubate, Ignite — the center for clinical research innovation, established by DCRI to drive meaningful improvements in the speed, efficiency, and impact of clinical research.
Project Loom
A flagship initiative launched by i-Cubed to transform how clinical trials are run, using AI to automate the most time-consuming and error-prone parts of the process.
Before a single patient can be enrolled in a clinical trial, research teams must navigate a maze of document generation, regulatory filings, IRB submissions, and protocol development.
6-9 months
Clinical trial setup routinely took 6-9 months before a single patient could be enrolled.
75+ pages
Trial protocols could run 75+ pages, requiring extensive manual drafting and revision.
Manual workflows
ClinicalTrials.gov submissions and IRB workflows were largely manual and error-prone.
No compliance infra
Existing AI tools lacked the compliance infrastructure required by academic medical centers.
Project Loom combined Health Universe's AI infrastructure and multi-agent orchestration system with DCRI's deep expertise in trial design, regulatory requirements, and research operations. Within weeks, Project Loom developed AI-enabled workflows capable of supporting key trial activities.
12+ specialized agents
Working as one coordinated system to handle distinct phases of clinical trial setup.
Protocol generation
Brief synopses converted to full 75-page protocols automatically.
Regulatory submissions
Automated ClinicalTrials.gov submissions integrated end-to-end.
Human-in-the-loop
Full audit trails, inspectable outputs, and human-in-the-loop controls throughout.
Academic medicine and clinical research face uniquely demanding compliance requirements. Having regulatory-grade infrastructure was a key factor in DCRI's decision to partner with Health Universe.
ONC B.11 Certified
Health IT certification for interoperability and data exchange standards.
SOC 2 Type II
Independent audit of security controls, availability, and confidentiality.
HIPAA Compliant
Full compliance with healthcare data privacy and security requirements.
TEFCA Participant
Trusted Exchange Framework and Common Agreement for nationwide health data exchange.
These certifications — which often take competitors 12-18 months and significant capital to achieve — are built natively into Health Universe's technology, enabling rapid deployment.
Project Loom exceeded expectations across every measured dimension. Results were evaluated against a real clinical trial with rigorous methodology that gave DCRI confidence in scientific integrity.
| Metric | Before | After (Project Loom) |
|---|---|---|
| Trial setup timeline | 6-9 months | ~7.5 days (30-40× faster) |
| Document processing speed | Baseline | 10× faster |
| Time savings across key tasks | Baseline | 93% reduction |
| Protocol generation | Manual drafting (weeks) | Automated from synopsis |
| Regulatory submissions | Manual, fragmented | Automated, integrated |
| Audit trail | Inconsistent | Full, inspectable, traceable |
Rather than spend months on paperwork using manual and error-prone processes, researchers can drastically accelerate clinical trials setup.
Every day saved in clinical trial setup is not just money saved. It's the difference between a patient who can access an experimental therapy and one who cannot, or a patient who receives a lifesaving treatment in time and one who continues to wait.
Health Universe brings the platform discipline needed to generate, test, and improve AI agents for clinical trials.
— Christoph Hornik, MD, MPH, Principal Investigator, DCRI
©2026 Health Universe, Inc. Confidential. All rights reserved.