The patient-owned layer that makes a fragmented Australian health system behave like a single, reasoned, shareable record.
Platypus pulls AU Core FHIR data from every server a patient can reach, including the Sparked reference endpoints, into an encrypted vault on their device. It reconciles what it finds, resolves the codes, and shares on the patient's terms. Data held by companies becomes data owned by the patient.
STATUSPre-1.0 and pre-release, not in the app stores. One cross-platform React Native app, offline by design. In early testing on iPhone and Android, with Android as a direct APK rather than the Play Store. Sync is on-demand today.
Built in pairs
Every capability has a wire half and a human half. The FHIR verb is the means; the patient verb is the point.
Aggregate→Own
Records from every clinic, lab and specialist land in one vault the patient holds, on their phone, not on anyone's servers. The whole vault leaves with them as a standard Bundle, and their own entries are first-class records alongside it.
Request, Dispense and Statement collapse into one medication list with repeats derived, not guessed. Cross-source conflicts surface as field-level diffs. Codes become plain words, and the history reads as a timeline of episodes, not a flat list.
reconcile by AMT codeclinical-hash conflict diffNCTS $lookup + $expand
SNOMED CT-AU, AMT, LOINC
Share→Use
A summary the patient composes leaves as an encrypted link, a Bundle, or a readable PDF. It can be shown at the desk, sent ahead, or written back as a clearly labelled patient note with honest delivery status. Self-entered records travel opt-in, attributed via Provenance.
SHL create + receiveIPS and AU PS $summaryOID4VP check-in
Australia's FHIR effort today, across vendors, GP systems, clinics and the Sparked program, is building the server side well: conformant endpoints and reliable exchange between systems. That is half the battle.
The other half is the patient experience and the outcomes it drives. When people can view, understand and update their own health data, care gets better: an accurate medication list at every appointment, safer prescribing conversations, questions asked earlier. Platypus is that half. And because it uses AU Core the way patients actually will, it returns concrete findings like these.
WHAT THE PATIENT SEES
💊
Atorvastatin 40 mg
One entry, reconciled from two places, 2 repeats left
Taking now
GP clinicCity cardiologist
↓the same records, read the other way
WHAT THE ECOSYSTEM LEARNSfed back to the IG
GP clinic sends
"medicationCodeableConcept": {
"coding": [{ "system": "sct", … }]
}
City cardiologist sends
"medicationReference": {
"reference": "#atorvastatin-40"
}
Same prescription, two equally legal shapes. Pre-matched vendor pairs never hit this; a patient app reading both servers did, and the finding went back to the IG.
Shipping today
CapabilityWhat shipsGrounding
Multi-server pull + reconciliationSMART App Launch (standalone, PKCE, public client) against AU Core servers including the Sparked reference endpoints. Cross-source conflict detection and reconciliation that no single server can perform.hl7.org.au/fhir/core
Encrypted vaultSQLCipher AES-256 with the key in the Secure Enclave. Biometric unlock, lock-on-background, encrypted backup with a recovery code. Full-vault export as a standard FHIR Bundle. Works with zero server connections. Security-reviewed.on device
Validation against real endpointsFast structural pre-checks on the device, then the conformance verdict from $validate against a configurable AU Core validation server. The Sparked reference server by default, or the endpoint you actually intend to send to. AU Core 1.0.0 and 2.0.0, with server vendor badges.$validate
Terminology, resolvedNCTS Ontoserver $lookup, $expand and $validate-code over SNOMED CT-AU, AMT and LOINC. Plain language the patient can read, never guessed by a model.NCTS
Share, both directionsSmart Health Links created and received (the decryption key travels in the URL fragment), SMART Health Card import via QR, and IPS or AU PS Bundles composed from the vault, exported as an encrypted link, a FHIR Bundle file, or a readable PDF.hl7.org/fhir/uv/ips
Check-in and writebackProvider-initiated share to a clinic (OID4VP, same-device and kiosk QR). Try the live clinic verifier that works with Platypus today. Plus capability-gated patient-note writeback (DocumentReference, always attributed to the patient, never disguised as clinical data) following the US Core clinical notes pattern, with honest delivery status.live demo ↗hl7.org/fhir/us/core
Conformance, checked against the real thing
Not "we follow the standards", and not a verdict invented on the phone. Platypus asks a real AU Core server whether a record conforms, so you know it will be accepted before it moves.
01
Structural pre-checks, on device
Fast, offline checks of shape and required fields catch the obvious problems before anything leaves the phone.
02
The verdict from $validate
The IG-conformance verdict comes from a $validate call to a configurable AU Core validation server. The Sparked reference server is the default.
03
Point it at your own endpoint
Validate against the GP or specialist server the record is actually headed for, so acceptance is known, not hoped.
Most health software holds data about you, on someone else's servers. Platypus inverts that. The record lives with the patient, in a secure, private, transferable form, and the patient is the only party holding all of it.
Offline by design. No data harvesting, no collection. The vault works with zero server connections.
One cross-platform React Native app, built so the record is never tied to a single platform. iOS is the current focus; Android is planned.
The entire vault exports as a standard, encrypted FHIR Bundle. Nothing locked in, nothing tied to one device.
The data travels with the patient, between providers and between platforms, never locked down to either.
Roadmap
The road ahead
Today Platypus connects to AU Core FHIR servers, including the Sparked reference endpoints. The aggregation, reasoning and plain-language layer it builds is exactly what would make My Health Record usable for patients. That is where it is heading, once integration is possible.
My Health Record integrationBackground + auto syncNew-record notificationsTerminology provenance badges
See it before release
Platypus is pre-release. If you want early access, a walkthrough, or to point it at your AU Core server, get in touch.
Platypus
