Reviewer Guide¶
This guide routes reviewers to the right deeper read in about 90 seconds. It does not replace the product narrative in Portfolio Case Study.
90-second read¶
Scribe IQ is a governed clinical documentation AI prototype built on synthetic data only. It demonstrates:
- an offline synthetic clinical corpus pipeline (Corpus Artifacts),
- a Postgres/pgvector serving layer (System Overview),
- provider-agnostic LLM and embedding integrations (Provider Guide),
- clinical documentation workflows (Portfolio Case Study),
- Responsible AI audit surfaces (Privacy and Provider Boundaries).
The project does not claim PHI readiness or production clinical deployment.
What to look for¶
| Signal | Where |
|---|---|
| Healthcare product thinking | Patient chart, encounter viewer, pre-meeting prep, note generation in Portfolio Case Study |
| Data platform thinking | data_prep/, generated corpus artifact, dataset card, audit report in Corpus Artifacts |
| AI engineering | Grounded RAG, embeddings, provider abstractions in System Overview |
| Responsible AI | ai_interactions, redaction, citations, audit dashboard in Privacy and Provider Boundaries |
| Production judgment | Deferred SSO, tenancy, BAA, PHI controls in Product Context |
Education-to-healthcare bridge¶
My background is in governed education data platforms: longitudinal student records, advising notes, privacy-sensitive analytics, and AI decision support. Scribe IQ translates that architecture into a healthcare-shaped system: longitudinal patient records, clinical notes, grounded retrieval, auditability, and human review boundaries. The deeper narrative is in Portfolio Case Study.
Role alignment¶
Scribe IQ is one portfolio artifact among several. For how this repo maps onto role patterns across academic health, university IT, research, education innovation, and AI platform architecture reviews — including the architecture claims and evidence pointers — see Target Role Alignment.
Suggested review paths¶
Recruiter / hiring manager¶
Technical architect¶
Data platform reviewer¶
What this is not¶
- Not a production clinical system.
- Not PHI-ready.
- Not a packaged SaaS app.
- Not just a chatbot.
- Not dependent on real patient data.