Transparent, local-first visualization of longitudinal laboratory data using approved synthetic patients.
OpenAI Build Week 2026 demonstration
LabDelta explores how laboratory reports from different points in time can be transformed into transparent, traceable, and clinically reviewable longitudinal patient views.
The current repository contains the contest demonstration presented during OpenAI Build Week 2026.
- Dashboard overview across multiple patients
- Patient detail comparison
- Longitudinal laboratory trends
- Laboratory profile grouping
- Mathematical change visualization
- Provenance tracking
- Original laboratory report viewer
- Local-first architecture
- Demo walkthrough in multiple languages
This repository demonstrates data representation and transparency.
The current version:
- uses only approved synthetic patients
- is intended for demonstration and research
- has not undergone clinical validation
- does not provide diagnoses
- does not recommend therapies
- does not prioritize laboratory findings
- does not replace clinical judgement
- Rust
- Tauri
- React
- TypeScript
- SQLite
This repository represents the OpenAI Build Week 2026 contest demonstration. Future development will focus on transparent longitudinal laboratory visualization and collaboration with clinical partners for medical validation.