ExpertBoard is an experimental, standalone prototype for AI-assisted complex case review.
It explores reusable expert boards organized by gene, disease, or clinical domain. Each board can open case-specific review rooms where multidisciplinary experts review evidence, record positions, track consensus, and keep final decisions human-governed.
- Install Docker Desktop or Docker Engine.
- Start Docker locally before running the development environment.
- Make sure ports
8010and13307are available, or change them in.env.
git clone https://github.com/PubCaseFinder/expertboard.git
cd expertboard
cp .env.example .env
docker compose up -d --buildOpen:
http://localhost:8010/boards
Create the first demo board room from the landing page.
- Expert board list and demo board creation
- Case review rooms under a fixed expert board
- Review room with left-side workspace navigation
- Core board roles and optional support roles
- Phenotype, variant, hypothesis, evidence, history, and briefing views
- MySQL-backed case, member, comment, decision, and audit log tables
- PubCaseFinder API client placeholder via
PUBCASEFINDER_BASE_URL
Core board:
- Case chair
- Clinical reviewer
- Bioinformatician
- Laboratory scientist
- Genetic counselor
Optional support:
- Case coordinator
- External consultant