ClinicalTrial OS

ClinicalTrial OS is an open-source AI research tool that helps clinical trial planners and healthcare researchers analyze the trial landscape and pick sites using public data from ClinicalTrials.gov and the FDA. You search by drug, sponsor, condition, or NCT ID and open a full trial detail view, then the platform splits eligibility criteria into inclusion and exclusion lists and scores each trial's complexity against peer trials. It ranks real trial sites by their experience, completion rate, and current activity, runs an early-stop risk model trained on finished trials, and applies a 23-point registry quality and transparency check. A knowledge-graph view, recruitment prediction, and adverse-event tracking round out the analysis. The app is built in Python with Streamlit, SQLite, pandas, scikit-learn, NetworkX, Plotly, and ReportLab, and is released under an MIT license; optional AI features need a free Groq API key or a local Ollama setup. The hosted demo covers 4,785 trials across 10 disease areas and 32,498 sites worldwide. The project states it is for research and educational use only, not regulatory or medical advice. For teams weighing the best AI for clinical trial feasibility work, BestAIFor lists it as a free, self-hostable option among AI research tools.

Clinical trial feasibility, eligibility complexity scoring, and site selection from ClinicalTrials.gov and FDA data