README.md

Documentation index

Canonical entry point for Scribe IQ documentation. Start with the overview section if this is your first visit. For a two-minute what runs now spine that points at the baseline, see architecture/CURRENT.md.

---

If you read only three docs

1. ../README.md — product framing, screenshots, architecture sketch, and quickstart.
2. overview/REVIEWER_GUIDE.md — fastest path through the repo by reviewer type.
3. overview/PORTFOLIO_CASE_STUDY.md — product, architecture, governance, and production-boundary narrative.

---

Overview (read first)

Doc	What it is for
overview/REVIEWER_GUIDE.md	Fast review paths for recruiters, technical architects, and data platform reviewers
overview/PORTFOLIO_CASE_STUDY.md	Product case study: problem, workflows, architecture, governance boundaries, and production limits
overview/TARGET_ROLE_ALIGNMENT.md	Role-fit interpretation for academic health, university IT, research, education innovation, and AI platform architecture reviews
overview/PRODUCT_CONTEXT.md	Problem framing, scope, what is deferred and why
overview/SYSTEM_OVERVIEW.md	Diagrams, capability flags, extension seams (rationale in DESIGN_NOTES)
overview/DESIGN_NOTES.md	Builder's perspective: alternatives considered, what was non-obvious, what would change for production
overview/PRIVACY_AND_PROVIDER_BOUNDARIES.md	Synthetic data boundary, provider egress, audit redaction, and PHI caveats

Run and operate

Doc	What it is for
guides/CORPUS_ARTIFACTS.md	Generated corpus artifact contract and missing-data expectations
guides/QUICKSTART.md	One supported path to a working local system
guides/LLM_AND_EMBEDDING_PROVIDERS.md	LLM and embedding provider setup (Groq, Azure OpenAI, Bedrock)
guides/README.md	Contributor setup and conventions

As-built reference

Doc	What it is for
architecture/IMPLEMENTED_BASELINE.md	Authoritative inventory: routes, schema, env flags
architecture/CURRENT.md	Short narrative of what runs today
architecture/README.md	Architecture hub with pillar links

Per-pillar READMEs

Pillar	Location
Backend (FastAPI)	backend/README.md
Frontend (Next.js)	frontend/README.md
Offline corpus pipeline	data_prep/README.md

Corpus pipeline reference

• reference/corpus_offline_pipeline_v2_brief.md — execution source of truth for scripts 01–09
• reference/data_prep_adapt_notes_longitudinal_design.md — adapt notes plus longitudinal context contract

Roadmaps

Doc	Scope
roadmap/SCRIBE_IQ_UI_ROADMAP.md	UI roadmap (forward-looking)

When a roadmap disagrees with architecture/IMPLEMENTED_BASELINE.md, trust the baseline until the roadmap is updated.

Project owner utilities

Doc	What it is for
overview/RESUME_SNIPPETS.md	Reusable project descriptions for external profiles and application materials

Archived design lineage

Doc	What it is for
archive/README.md	Superseded roadmaps, design drafts, and hygiene checklists with successor pointers

Visual assets

• README and social showcase images: assets/showcase/readme/ and assets/showcase/social/og.png
• UI mockups and screenshots: design/README.md

Lineage and history

• history/EVOLUTION.md — short timeline of how the layout and docs evolved
• archive/README.md — superseded prompts and drafts kept for record
• Archived precursor scripts: corpus_pipelines/agbonnet_hf_clinical_notes/ — preserved for reference; do not extend

Generated artifacts

Output of the corpus pipeline, not hand-maintained specs. The data/clinical_corpus_v2/ JSONL files are generated/restored corpus artifacts, not committed application source; see guides/CORPUS_ARTIFACTS.md.

• data/clinical_corpus_v2/patients.jsonl
• data/clinical_corpus_v2/encounters.jsonl
• data/clinical_corpus_v2/notes.jsonl
• data/clinical_corpus_v2/manifest.json
• data/clinical_corpus_v2/dataset_card.md
• data/clinical_corpus_v2/audit_report.md