Geometric lines of the human body

The Deep-Tech Partner for AI, Simulation & Robotics

Generative AI + RAG Clinical Trial Optimizer & Protocol Controller

Woody de Kafou | February 23, 2026

Med-techGenerative AIMachine Learning

The Problem

Clinical trials are slowed down and made more expensive by:

  • Overly complex protocols that are hard to recruit and retain for
  • Slow, manual amendment cycles
  • Site teams wrestling with dense PDFs and inconsistent guidance
  • Risky deviations whenever real life collides with rigid plans

Sponsors and CROs need a system that continuously optimizes trial design and execution—not just another static protocol document.


The Solution: A “Clinical Trial Brain” Powered by GenAI + RAG

Propel builds a Generative AI + Retrieval-Augmented Generation (RAG) platform that sits across your clinical trial lifecycle:

  • During Design – Trial Optimizer
    • Drafts and refines protocols using your historical trials, regulatory guidance, and internal SOPs
    • Scores complexity and burden for patients, sites, and operations
    • Forecasts recruitment and feasibility, and compares “what-if” protocol variants
    • Sandboxes amendments, showing impact on timelines, burden, and risk
  • During Execution – Protocol Controller
    • Acts as a natural-language copilot for sites: “What’s allowed if this visit is late?”
    • Retrieves the exact protocol and guidance text behind every answer
    • Monitors deviations and patterns across sites in near real time
    • Suggests compliant, low-risk operational adjustments within defined guardrails

All interactions are traceable, auditable, and grounded in your real documents and data.


How It Works (At a Glance)

  1. Ingest & Index
    • Protocols, amendments, prior trials, regulatory guidance, SOPs, site manuals, and operational data are securely ingested and indexed.
  2. Reason with GenAI + RAG
    • GenAI handles drafting, explanation, and scenario analysis.
    • RAG keeps every recommendation tied to specific source texts and data.
  3. Close the Loop
    • Execution data feeds back into the Optimizer, improving future protocol design, feasibility estimates, and operational playbooks.


Business Value

  • Faster, higher-quality protocol designFewer amendments, smoother start-up, and more feasible trials from day one.
  • Better recruitment & retentionProtocols optimized for patient and site burden, with real-time operational support.
  • Fewer deviations, stronger complianceSites get instant, grounded answers; sponsors get full traceability and audit readiness.
  • A learning system, not a one-off toolEvery trial makes the “Clinical Trial Brain” smarter for the next.


Who It’s For

  • Biopharma sponsors running multi-site trials
  • CROs wanting differentiated, tech-enabled services
  • Healthcare / life sciences teams on AWS looking to operationalize GenAI safely

30 minute technical discovery

Woody de Kafou

Woody de Kafou

Founder & CEO, AI Thought Leader

BOOK A MEETING