RePORT AI Portal

RePORT AI Portal is a local-first assistant for one clinical research study. It helps a study team load local source files, publish a PHI-scrubbed study bundle, and ask grounded questions from that bundle.

This README is only the front door. The full user, IRB/auditor, and developer documentation lives in Sphinx: https://solomonsjoseph.github.io/RePORT-AI-Portal/

Start Here

Need	Go to
Use the portal	User guide
Run the first study	Quick start
Configure models and study settings	Configuration
Review PHI handling	IRB/Auditor profile
Change or maintain code	Developer guide

Quick Start

Install uv for your platform:

# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# Windows PowerShell
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Then start the portal:

git clone https://github.com/solomonsjoseph/RePORT-AI-Portal.git
cd RePORT-AI-Portal
make chat

make chat installs the web/AI dependencies it needs, launches the web UI, and guides provider selection, study loading, PHI-key setup, and chat. For the complete setup path, use the installation and quick start pages.

Privacy Boundary

Treat source study files as PHI-bearing unless your study team has verified otherwise. The assistant is designed to answer from the published scrubbed bundle, not from raw source files. IRB, IEC, privacy, and audit reviewers should start with the IRB/Auditor profile.

Development

Use the developer guide for architecture, testing, operations, production readiness, and contribution workflow.

Common local checks:

make test
make docs-quality

Support

Open an issue for bugs or documentation gaps: https://github.com/solomonsjoseph/RePORT-AI-Portal/issues