The HEMI way: for whom the pager tolls

In an interview with HealthTechAsia, Dr. Ezam Mat Ali, CEO and Founder of Health Medical Intelligence (HEMI) in the UK and MedPlanner in Malaysia, explains how his three-product AI platform is designed to close the gaps that hold healthcare systems back — on both sides of the world.

A paediatrician who spent 16 years practising in the NHS and 29 years in the UK in total, Dr Ezam Mat Ali returned to Malaysia in January 2024 to relaunch MedPlanner, his AI health technology company, with a new founding team, with a clear mission: to build technology that addresses some of the biggest structural challenges facing hospitals, clinicians, and patients — and then take it back to a global market. 

We met at a place very familiar to me: the UOA building in Bangsar, just a few minutes from where I used to live in KL, and from where I used to take the LRT daily. Life is fun like that, constantly throwing up reminders of the past.

Another reminder came in the form of his London business address in Covent Garden, where my then-employer’s office was located, ironically, before I came out to Malaysia in the first place, many years ago.

Over the next hour, he introduced MedPlanner’s product suite: three interconnected solutions built around a proprietary clinical AI engine called HEMI (Health Medical Intelligence), designed to serve the full healthcare ecosystem: clinical teams, individual clinicians, and patients.

“You have the hospital and the clinical teams. You have the individual clinician and nurse. You have the patient and the public,” Dr. Ezam said. “These three layers need to be connected, and the information and intelligence need to flow across them. That is what the three products together are designed to do.”

His years inside the NHS give him an unusually clear view of why that matters. Waiting times are long not simply because demand is high, but because the system haemorrhages time, to fragmented communication, to documentation done retrospectively, to patients who arrive underprepared and leave underinformed. “I know what busy hospitals look like from the inside. That is the context everything here is built from.”

Malaysia presents a different but equally familiar set of pressures: high rates of non-communicable diseases, stark disparities in health literacy, and a mixed public-private system that makes coordination across care settings genuinely difficult. “I have been in both systems. These are not abstract problems to me.”

Both markets, he believes, are ready, and both are part of the launch strategy.

“The UK is our global headquarters for distribution. The reason I came back to Malaysia was to build these products faster and at lower cost than I could in the UK. But the target markets are global — NHS, Ministry of Health Malaysia, and international health systems.”

MedPlanner completes clinical AI ecosystem with launch of HEMI Health and AskHEMI

Connecting clinicians

The three products are designed to work together, each targeting a different stakeholder in the same ecosystem. The first, HEMI Teams, is a clinical coordination platform.

“It was actually the first product I built, before I came to Malaysia. The original version shared the company name: MedPlanner. The insight behind it came from watching junior doctors in my department using WhatsApp to manage patient care. They knew they weren’t supposed to. But there was nothing on the ground that worked.”

He cited oncology as an illustration. In both the NHS and Malaysia’s private sector, cancer care coordination remains a persistent challenge. When six or seven specialists are involved in a single patient’s care, the question of how they communicate — how information is transferred, shared, and discussed securely — has no clean answer in most clinical settings.

The problem compounds when two separate clinical teams need to speak to each other. On WhatsApp, the only workaround is to create a third group containing everyone, which quickly becomes unworkable. “HEMI Teams is designed so that one team can invite another into a patient case, creating a shared channel, while each team retains its own private discussion space. Nobody has built this for healthcare before.”

The platform’s foundation is clinical workflow rather than AI. Getting that right comes first; HEMI then supports efficiency and intelligence on top of it. All case data is captured and retained, and a clinician can ask HEMI directly, mid-discussion, what might be missing from a care plan, or what else should be considered. It synthesises everything discussed and surfaces clinical recommendations without replacing clinical judgement.

Over time, that accumulated data becomes an asset in itself. A thematic analysis function allows HEMI to identify patterns across a population of cases: the most discussed conditions, the most frequently prescribed medications, trends in clinical decision-making. It is the kind of intelligence that only becomes visible at scale.

Bringing back ‘pyjama time’

While HEMI Teams works for groups, HEMI Health is for individual clinicians: an AI medical scribe and clinical intelligence tool.

“If you ask any doctor or nurse what the most tedious part of their job is, 95 per cent will say documentation,” Dr. Ezam said. “Clinicians spend, on average, four to six minutes writing notes for each patient after a consultation. If you see 30 patients, that is two hours of what we call ‘pyjama time’ — staying late or coming in early to complete documentation that should have been done in real time.”

The administration burden does not end there. A nurse or junior doctor handing over at the end of a shift can have up to 30 patients to work through: walking each incoming colleague through current status, pending actions, and what has been done. That alone can take an hour. On a heavy day, a clinician could easily spend three hours on documentation and handover combined.

HEMI Health addresses both. The clinician records the session — whether a doctor-patient consultation, a doctor-to-doctor discussion, a dictation, or an MDT meeting — and HEMI transcribes and structures it into medical notes: clinical impression, differential diagnoses, recommended investigations, and a handover summary. The clinician reviews and confirms rather than having to recall and reconstruct. The shift, as Dr. Ezam puts it, is from active documentation to verification.

What distinguishes HEMI Health from rival AI scribe products, according to its founder, is its integration with the broader HEMI ecosystem — feeding directly into HEMI Teams and generating clinical letters, referral notes, and handover summaries from the same session — and its foundation in medico-legal principles.

“The clinical reasoning element comes from my work with the Medical Protection Society in the UK, where I was a medical advisor on patient complaints and medico-legal cases. I learned what protects doctors when things go wrong. A lot of the time, a doctor has done everything right — but if the reasoning behind the decisions is not documented, it becomes very difficult to defend. If it is not documented, it did not happen. That is how it is treated legally.”

HEMI Health is designed to close that gap by generating a record of clinical reasoning at the point of consultation. “Even if outcomes are negative, the documentation shows what was known and considered at the time. That is what makes it defensible — and what sets it apart.”

Underpinning all three products is a clinical evidence engine — a searchable knowledge base spanning more than 200 million peer-reviewed papers, with outputs structured by hierarchy of evidence: meta-analyses, systematic reviews, and randomised controlled trials. A clinician can submit a query and receive a synthesised evidence summary with full citations linking directly to the original studies, filterable by journal tier for those who want outputs drawn only from the highest-ranked publications.

“If a clinical decision is backed by published evidence, it becomes very difficult for anyone — a patient, a lawyer, a regulatory body — to challenge it.”

News at Ten

Available on iOS and Android with a free tier, AskHEMI is designed to support patients between clinical visits.

“After a consultation, particularly for a new diagnosis or a complex condition, how much does the patient actually retain?” Dr. Ezam asked. “In my experience, not much. They may remember the tone of it, how it felt, but the clinical detail fades quickly.”

AskHEMI is there for the moment at 10pm when the patient remembers they had a question, or when a test result arrives and they do not know what it means. It is not a replacement for the doctor. It fills the gap between appointments.

Where a general-purpose AI tool provides broad answers, AskHEMI draws on the same clinical evidence engine that powers the broader HEMI ecosystem, directing that depth specifically at patients and the public. It is also multilingual, covering Bahasa Malaysia, Mandarin, major UK regional accents, and a range of local dialects. “Malaysia is actually an ideal environment for developing and testing something like this — the linguistic diversity here is genuinely useful.”

Location intelligence is built in. If a user is managing a condition and needs to see a specialist, AskHEMI can surface nearby clinics and practitioners with contact details. Users can also upload photographs for skin lesion analysis — receiving likely classifications, an assessment of concern, and guidance on whether to seek professional review — or submit blood test reports and imaging results for a contextualised explanation.

The premium tier, at around ten ringgit per month, adds a personalised health profile. Users enter their biometric data — blood type, cholesterol levels, resting heart rate, and more — and can upload medical documents. Questions are then answered in the context of their specific profile rather than in the abstract.

“The intention is not to replace a consultation,” Dr. Ezam said. “It is to make the patient a more informed participant in their own care, before they walk in, and after they walk out.”

It is a philosophy, he says, that runs through everything MedPlanner has built.

“The best outcome for any government, for any health system, is prevention: keeping people from becoming patients in the first place.”