System dossier · Entry 007
MediAssist
HEALTHCARE · DECISION SUPPORT · 2025
Status
Prototype · Research
Scope
Healthcare decision support
Operational summary
MediAssist explores voice-first symptom intake combined with classical ML models for preliminary analysis — always framed as assistive, not diagnostic.
The pipeline captures spoken input, structures symptoms, runs prediction models trained on curated datasets, and generates readable reports for review.
Interface capture pending · Ref. 007
Fig. 007 — Interface capture
Technology stack
Frontend
- React
- Web Speech API
Backend
- Python
- Flask
- scikit-learn
Data & ML
- SQLite
- CSV training sets
Infrastructure
- Local / cloud deploy
Operational pulse
Voice-first intake — assistive, never diagnostic.
Voice
Symptom capture
ML
Classical models
Report
Readable output
Module 01
Speech input
Spoken symptoms structured automatically.
Operational unit→
Module 02
Analysis
Traditional ML on curated datasets.
Operational unit→
Module 03
Prediction
Preliminary signals for review.
Operational unit→
Module 04
Report gen
Human-readable summaries exported.
Operational unit→