MedVita AI platform for hospital operations

The problem to solve

Vinmec Da Nang already uses electronic medical records and operates to JCI standards, yet pre-visit history-taking, in-visit documentation, record completion and follow-up after patients go home still take much of clinical staff time. The project aims to reduce documentation and data-entry time, shorten waiting times, keep records complete and consistent, and support continuous follow-up after visits, surgery and discharge.

Four proposed solutions

The proposal comprises four solutions sharing data along the care journey: before the visit → during the visit → after the visit → between visits.

• M1 – Smart patient intake (before the visit): patients answer an AI interview through existing booking channels or a kiosk; international patients are interviewed in English and doctors receive a Vietnamese summary; red-flag signs are escalated to nurses at once. Nurses review before anything enters the record.
• M2 – Clinical documentation (during the visit): structured notes covering symptoms, assessment, orders and treatment plan. Recording only with patient consent; the doctor checks, edits and confirms.
• M3 – EMR automation (after the visit): confirmed data is compiled to pre-fill the record and flag missing mandatory fields; the doctor approves before it is transferred to the EMR.
• M4 – Personalised care (between visits): AI drafts home-care plans, medication and follow-up reminders, care instructions and exercises; it checks in periodically and flags warning signs early. Applied only after clinical staff approval.

Role of technology

According to Trivita AI’s proposal, MedVita connects to the existing EMR rather than replacing hospital systems, and can run standalone while integration is completed. These features are as described by the vendor; scope of use at Vinmec Da Nang and accuracy in Vietnamese will be confirmed through appraisal and testing.

Role of the hospital

The hospital selects pilot departments, appoints clinical and IT leads, approves workflows and retains clinical oversight. Decisions on integration, data security and scale-up involve the Vinmec Healthcare System. AI does not replace doctors’ diagnosis, orders or advice.

Data and safety

Data-processing scope and security requirements are agreed in line with hospital policy, Vinmec Healthcare System rules and personal data protection law. The hospital formally informs patients and staff before deployment; visits are recorded only with patient consent.

Proposed timeline

Per the proposal, the project runs six months (November 2026–April 2027), introducing solutions in three waves; each wave is trialled in one or two departments chosen by the hospital and adopted only after meeting acceptance criteria. Milestones apply only if the proposal is approved and may change; contract signing, acceptance and first patient service are separate milestones.

• Preparation and integration (process review, design, HIS/EMR connection): November–December 2026.
• Wave A · M1 Smart intake: preparation → trial → go-live 15 February 2027.
• Wave B · M2 Clinical documentation, M3 EMR automation; Wave C · M4 Personalised care: preparation → trial → go-live 1 April 2027.
• Proposed acceptance milestones: 20 Nov 2026 scope sign-off · 1 Dec 2026 design approval · 30 Jan 2027 pilot approval · 28 Apr 2027 overall acceptance.
• Impact evaluation: baseline measured before the project; outcomes measured during use.

Monitoring and evaluation

Indicators must be agreed by the team before measurement and compared with the pre-project baseline. No outcome data is published here.

• Documentation and record-completion time; patient waiting time.
• Share of AI content requiring edits; completeness and consistency of records.
• Clinical staff acceptance; experience of domestic and international patients.
• Safety, data incidents and feasibility of scale-up.

About the proposing company

Summary of the MedVita 2026 company profile supplied by Trivita AI. Information is self-reported by the company and will be appraised separately by the hospital.

• Trivita AI: researches, develops and deploys AI solutions for sector-specific problems; incorporated in Vietnam in 2025; the founding team has worked on AI in healthcare since 2021.
• Ecosystem: MedVita (hospital operations), VitaLaw (law), EdVita (education).
• Leadership: Prof. Ho Pham Minh Nhat, Founder & CEO (Department of Statistics & Data Sciences, University of Texas at Austin); Prof. Nguyen Quang Phuoc, CPO (University of Ulsan, Korea).
• Experience cited: an AI system for Thai Nguyen National Hospital (2026); AI support in mental health care in Korea (2022).
• Technology approach: in-house core algorithms and a Vietnamese medical large language model; data stored in Vietnam; integration with HIS, EMR and PACS.

Status on 1 November 2026

The project has not been approved and the proposal is under evaluation, focusing on technology and data security. No contract has been signed and nothing has been deployed. Next steps proceed only with approval from the hospital and the Vinmec Healthcare System.

• Trivita AI proposal received: October 2026.
• Technology and security evaluation: in progress.
• Approval: not yet granted; to be updated separately according to the hospital’s decision.
• Pilot: not started.