φ(ai) PHI AI
DocumentationDocs
Repository
PHI AI
The AI-native platform for protected health data

True AI, securely and compliantly,
running in your own cloud.

Every practice and hospital sits on decades of records an AI could read in seconds — and almost none can safely let it. The lock is there for good reason; it is also why care still runs on typing, faxes and hold music. PHI AI opens it without breaking it: a frontier model working on your real charts, in your own cloud, on your own keys. Notes written. Scans read. Denials fought. Questions answered. It sees only what your role allows. Compliance at its core, and purpose-built for AI. Suitable for family practices or large health systems.

We welcome feedback, questions, and criticisms — get in touch at www.ryangomez.nyc
Capabilities

What it does, out of the box

Every feature is AI-native. Everything below is running on this site right now, each powered by a managed model in the platform's registry — with live analytics, drift recognition and alerting behind every one.

PHI AI assistant

Retrieval-tuned AI over the record with citations per claim, role-scoped tools, and refusals that name their rule — on the foundation model you bring.

Denial risk prediction

Every pending claim scored before submission, from your own adjudication history — each factor cited to real counts.

Prior-auth evidence assembly

Criteria checked against the chart, evidence gathered, and the appeal drafted — into the signature queue, never past it.

Coding integrity

Two-directional review: findings the record supports but the codes miss, and billed codes the documentation does not support.

Release-of-information validation

Every release checked against the requesting jurisdiction's requirements; hard failures block with the requirement named.

Sensitivity classification

Records classified at the door — Part 2, mental health, HIV, reproductive, genetic — with low-confidence calls queued for humans.

Model library — bring your own

Any model, any kind: registered, enabled and activated into capability slots through one audited control panel.

Model analytics & drift

Live performance metrics, population-stability drift recognition, and alerts that never close themselves.

EMR integration, both directions

Six vendors, bulk and streaming, in and out — each vendor's real seams honored and refusals recorded.

Population analytics

Cohorts, quality measures, no-show risk, trial pre-screening — on a plane that returns counts, never patient lists.

Ambient documentation

Record a visit from the browser — behind a state-law consent gate — and the capture lands in the Data Store, on the chart, audited.

Imaging, read by vision AI

A genuine vision inference over the pixels: technique, findings, measurements from the acquisition scale — drafted to the queue.

Patient instructions, gated

Plain-language instructions written from the chart — then a second model pass fails the draft on a single invented assertion.

No-show risk & outreach

A transparent risk score — every factor and its points printed — and an AI call script per patient, staged for a human to dial.

Trial pre-screening

Protocols screened against the whole corpus in SQL; the model reads the charts as a recall-biased second reader that never rules out.

Quality measures, live

Real numerators over real denominators, computed from the corpus on page load — arithmetic in the open, counts never lists.

Fairness, screened

The claims model examined across sex, age and payer subgroups — calibration gaps, printed thresholds, an AI reading of the results.

Ingest & mapping QA

Integrity rules sweep every record in the store; the model argues root causes and fix order — it never quarantines anything itself.

Human signature, structurally

AI output is a draft until a licensed person signs it. Only the signed artifact acquires a write path to the EMR.

Audit, live

The hash chain is re-verified on screen, in front of you — including the refusals, which is the point.

Compatibility

Works with what you already run

Out of the box: the major EMRs in both directions, the three major clouds as your own account, and the frontier models — plus anything you bring.

Six EMRs, both directions
Epic
SMART Backend · Bulk FHIR
Oracle Health
explicit scopes · bulk
athenahealth
OAuth2 · bulk
eClinicalWorks
JWT assertion
MEDITECH
Expanse · g(10)
NextGen
per-patient reads
Three clouds — your account, your keys
Amazon Web Services
S3 · KMS · RDS · Bedrock
Google Cloud
GCS · Cloud KMS · Vertex AI
Microsoft Azure
Blob · Key Vault · Postgres
Any frontier model — and your own, under your own BAA
Claude
Anthropic API · Bedrock — this demo runs Sonnet 5
Gemini
via Vertex AI, in your project
GPT
via Azure AI Foundry, in your tenant
Llama
via Bedrock, in your account
Mistral
via Bedrock, in your account
Whisper
faster-whisper · self-hosted · MIT
Your model
any HTTPS endpoint · BYOM registry
Why it's different

Built for PHI. Not adapted to it.

The AI is inside the walls

The model reaches data only through role-scoped, audited retrieval tools — the same gates people go through. It receives excerpts, never a database connection, and every answer cites the records it came from.

Every read is a recorded disclosure

A hash-chained, append-only audit trail carries every question, every record read, and every refusal. Withheld records are counted and stated, never silently missing.

Compliance is the design, not a layer

Population tools return counts and cannot enumerate patients. Unsigned AI drafts have no write path. Psychotherapy notes never export. The refusal is in the shape of the API — not in a policy document.

Your cloud, your keys

Deploys into infrastructure you own: envelope encryption under your KMS, and a web tier that makes no third-party calls. You bring your own LLM — any foundation model, under your own BAA with your AI provider — and your organization's identity system governs who sees what.

Extensible

Any model of any kind joins the registry — foundation, predictive, classifier, imaging, or yours — behind HTTPS and an audited lifecycle.

Scalable

Two deployment shapes, chosen deliberately: a clinic on one profile, a health system at terabyte–petabyte scale on the other.

Configurable

Every switch, model, feed and threshold governed from the Control panel — and every change lands on the audit trail under a name.

Compliant by design

HIPAA minimum necessary, 42 CFR Part 2, psychotherapy separation — enforced in the architecture, verified live, documented in full.

Start here

See it working

Pick a role, ask the assistant a question, try to reach a screen your role cannot see — and then read the audit trail that watched you do it.