Hospital-grade AI, built in Malaysia
One patient. One record. Twelve stages that never break.
TruHealth AI carries a patient from the first search to long-term chronic care — appointments, check-in, consultation, ward, discharge, settlement and recovery at home — on a single platform where clinical AI, governance and interoperability are built in rather than bolted on.
The twelve-stage continuous journey
Most platforms model an episode of hospital care. This one models a relationship: stage 11 feeds back into stage 0. Select any stage.
Why this platform, and why now
Three shifts in the Malaysian market changed what a healthcare platform has to do. All three are addressed in the product, not just in the pitch.
DRG reimbursement reaches private hospitals
Payment moves to severity-based case groups, with a national system to follow. Documentation stops being paperwork and becomes the thing that determines what a case is paid.
MOH DRG roadmap; ProtectHealth briefingsof records carried a coding error
In a Malaysian teaching-hospital casemix study, most reviewed records were mis-coded and half of the re-grouped cases attracted a lower tariff. TruHealth Code catches this while the case is still open.
BMC Health Services Research, MY-DRG casemix studyhealthcare travellers in a single year
Malaysia's international patients follow a longer journey — remote enquiry, quotation, travel, treatment, cross-border follow-up. No competing platform treats it as a journey at all.
MHTC reported arrivals, 2025; MYMT 2026 campaignFigures are drawn from public sources compiled in the September 2026 market scan and should be verified against a primary reference before use in contract discussions.
Four suites on two planes
Suites are how the platform is sold — a hospital can start with one. The two planes underneath are not optional: no service holds patient data or calls a model except through them, which is what stops governance and interoperability being descoped when the schedule tightens.
TruHealth Journey
- Assist v2
- Book
- Record
- Check-in
- Queue
- Pay NEW
- Passport NEW
TruHealth Connect
- Ward+ v2
- Vitals
- Nurse Copilot v2
- Home NEW
TruHealth AI
- Scribe
- Code NEW
- Guide v2
- Vision NEW
- Rounds NEW
- Predict NEW
TruHealth Engage
- Community NEW
- Academy NEW
- Outreach NEW
- Sense
TruHealth Trust
Every AI suggestion carries a plain-language explanation and its evidence. Every clinician override is recorded to a tamper-evident ledger. Model cards, drift and bias monitoring, PDPA breach and portability workflows. Built on Truxity's TruGovern platform and SACI cryptographic identity.
TruHealth Exchange
FHIR R4 as the internal canonical model — not an export format. HL7 v2 and IEEE 11073 adapters, SMART on FHIR launch, CDS Hooks, MyHDD and MyHRDM mapping, and a governed MCP server so approved AI agents can act on hospital data under policy and audit.
The complete module register
Twenty-four modules across four suites and two planes. Marked entries are the ones no competitor scanned in September 2026 offers.
| Module | Layer | What it does | Status | Phase |
|---|
Three bets, not thirty features
Any single module can be copied in two quarters. These three are defensible because each is built on something Truxity already owns and a competitor would have to acquire. Each states the evidence still required before the claim is used commercially.
Governance is the platform
A healthtech competitor would need a governance platform, a cryptographic identity engine and a clinical product in the same company. TruGovern and SACI already exist inside Truxity, built for a separate government market.
Documentation that pays for itself
Global scribe vendors have the ambient engine but no MY-DRG grouper logic or Malaysian coder workflow. Local casemix vendors have the reverse. TruHealth Code is the only place both sit in one product.
The substrate other AI runs on
Retrofitting a FHIR-native canonical model and a governed agent tool surface onto a mature product is a rewrite, not a release. Adopting Exchange makes TruHealth the integration layer for whatever else the hospital buys.
Usability written as requirements
Malaysia's clinical workforce is short-staffed. A tool that costs thirty seconds a patient gets abandoned whatever its AI quality. So these are acceptance criteria with a named waiver authority, not aspirations — a module that misses its budget has not shipped.
- Zero-app front doorEvery patient action from discovery to check-in works over WhatsApp, SMS, a browser link or a phone call. App-only features are treated as design defects.
- The three-tap ruleBook, check in, see your queue position, pay a bill — three interactions or fewer, and always resumable without re-entering anything.
- Bilingual by constructionBahasa Malaysia and English are peers, and a single sentence can mix both, because that is how people actually write here.
- Designed for the least confident userAccessibility mode, caregiver proxy with recorded consent, and WCAG 2.2 AA as a build requirement rather than a remediation project.
- The clinician ten-second ruleNo AI output may cost more than ten seconds of review in the common case. A feature that cannot meet the budget ships switched off.
- Degrade gracefullyCheck-in, queue, vitals and documentation keep working offline for an hour and say plainly what is currently unavailable.
See it from your own seat
Seven roles, seven different definitions of a good day. Sign in as any of them — each screen shows only what that role is permitted to see, and why.
Open the demo