Researching AI for Safer, Smarter Clinical Trials
At Keiji AI, we conduct foundational research on AI models that support real-world clinical development — including evidence synthesis, trial design, patient simulation, and protocol generation.
Our research combines generative AI, real-world data modeling, and biomedical informatics to accelerate clinical research. We focus on practical, safety-aware, and regulatory-aligned systems for AI in trials.
Research Focus Areas
Automating systematic reviews, meta-analysis, and PRISMA-compliant workflows with LLMs.
AutoTrial and Panacea models to optimize eligibility criteria, endpoints, and arms.
Key Publication:
AutoTrial: Prompting Language Models for Clinical Trial DesignFoundation models that simulate digital twins, match EHRs to protocols, and forecast trial feasibility.
Key Publication:
Matching patients to clinical trials with large language modelsTools for validating AI performance on OMOP/CDISC datasets in real-world studies.
Our Research Principles
Scientific Rigor
We ground our models in peer-reviewed science and transparent benchmarking.
Clinical Relevance
Our work targets problems that matter — trial feasibility, patient access, and evidence generation.
Safety-Aware Modeling
We align to CDISC, PRISMA, and real-world regulatory standards in all workflows.
Open Collaboration
We work with academia, regulators, and pharma sponsors to ensure impact beyond papers.
Join Our Research Team
If you're passionate about advancing AI for health, we’d love to collaborate. Explore opportunities to contribute to cutting-edge research at Keiji AI.