How Thai AI Innovation Could Ease the Load on Our Healthcare System
This is a personal perspective. Not a policy paper. Just thoughts I keep having when I walk past a packed outpatient department at 8 a.m., or when a patient tells me they took a half day off work, queued for two hours, and got five minutes with a doctor.
Our public healthcare system in Thailand is, by international standards, remarkable. The people inside it work very hard. The pressure on them is also very real. Long queues, overworked staff, patients who travel far for routine medication refills — these are not small problems.
I think Thailand has a quiet opportunity here: to use AI and digital tools that we build ourselves, to take pressure off the system without removing the human care at its centre.
A few thoughts, from a clinician.
1. Telemedicine for follow-ups and stable conditions
Not every visit needs to happen in person. A diabetic patient whose numbers are stable, a dermatology follow-up, a post-procedure check-in, a mental-health touch point — many of these can be safely handled by a 15-minute video consultation.
What this changes:
- Patients don't lose a day of work or pay for transport
- Clinicians can focus the in-person slots on cases that truly need a physical exam
- Rural patients can reach specialists they'd otherwise wait months to see
Thailand already has the smartphone penetration. What we still need is integration with hospital systems and clear reimbursement so doctors are not penalised for choosing the format that's best for the patient.
2. Home delivery of routine medication
For patients on long-term medication — chronic disease, hypertension, common dermatology refills — the trip to the hospital is often only to pick up a prescription they've taken for years.
Pharmacy delivery, with proper cold-chain handling and patient verification, can:
- Free up pharmacy queues for patients who genuinely need counselling
- Reduce waiting-room congestion (and the infection risk that comes with it)
- Give elderly patients and caregivers their day back
The technology is straightforward. The harder part is the workflow — verifying identity, confirming adherence, and looping the prescribing doctor in if something changes.
3. AI-assisted triage to the right department
One of the most underappreciated bottlenecks in our hospitals is patients arriving at the wrong door. A symptom that should go to ENT ends up in general medicine; a child who needs paediatric surgery is first seen by general paediatrics; a complex skin issue is sent through three departments before reaching the right specialist.
A well-trained AI triage assistant — built on Thai-language symptom descriptions, Thai-population disease prevalence, and our actual hospital department maps — could help patients:
- Answer a short structured set of questions before they arrive
- Be guided to the right specialty on the first visit
- Skip duplicated check-ins and re-registration
To be clear: AI does not replace clinical judgement. It just helps the patient land in the right room, faster.
4. Secure inter-hospital data sharing
This is the one I feel most strongly about. Right now, a patient who moves from a community hospital to a tertiary centre often arrives with a paper file — or no file at all — and tells their history again. Lab work is repeated. Imaging is re-done. Time and money lost on both sides.
A standards-based, patient-consented, encrypted data-sharing layer between Thai hospitals would:
- Let receiving doctors see prior labs, imaging, and medications immediately
- Avoid duplicate tests
- Catch dangerous drug interactions earlier
- And — critically — keep the patient's information private and under their control
The technical pieces (FHIR-style data standards, end-to-end encryption, granular consent, audit logs) are well understood globally. The question for us is governance: who holds the keys, who audits access, and how patients revoke consent. These are solvable. They just need to be solved with patients in the room.
What I'd Hope Thai-Built AI Looks Like
If we are going to build these tools — and I hope we do, by Thai teams — a few things matter to me as a doctor:
- Privacy by default. Patient data is not a commodity.
- Trained on Thai data. Symptom phrasing, disease prevalence, drug names, even the way patients describe pain — all are local.
- Designed with frontline staff, not for them. The best digital tools I've ever used were shaped by the nurses and pharmacists who'd live with them daily.
- Augmenting, not replacing, the clinician. AI should give doctors more time with patients — not less.
A Quiet Closing
I am not a tech expert. I am a doctor who watches, every week, how much energy our system spends on logistics that could be quietly handled by good software, so that the human time can go to the parts only humans can do — listening, examining, deciding, comforting.
Thai engineers, Thai data scientists, and Thai clinicians together can build this. I hope we do.
— Dr. Waleerat (Dr. Gwang)
This is a personal essay reflecting Dr. Waleerat's own perspective as a clinician. It is not a policy statement, and not medical or technical advice.
