By Tony Low · AI in Action · October 23, 2026

Five years ago, doctors at EMC Healthcare’s hospitals in Indonesia were still working from paper charts. Today, the eight-hospital network is running one of the most complete AI deployments in Southeast Asian healthcare — spanning electronic records, radiology, ambient note-taking, retinal screening, and a chatbot that handles patient messages on WhatsApp. It’s a useful case study in what “AI adoption” actually looks like once you get past the pilot stage: not one flashy tool, but a steady accumulation of smaller ones, each solving a specific workflow problem.

The foundation was laid between December 2022 and June 2024, when EMC rolled out InterSystems’ TrakCare electronic medical record system across all eight of its hospitals, starting with its flagship Alam Sutera site. That 18-month migration from paper to digital records gave EMC’s roughly 700 specialist physicians and 100 general practitioners a single source of patient data for the first time — and set up the infrastructure for what came next.

In August 2025, EMC began piloting InterSystems IntelliCare, an AI layer built on top of that EMR, and says it became the first healthcare organization in Asia-Pacific to adopt it. Two features stand out. An AI assistant lets clinicians pull up patient histories, lab results, and medication lists through natural-language queries instead of clicking through multiple screens — InterSystems says it cuts the number of clicks needed to review a patient’s data from eight down to two. A second feature, ambient listening, transcribes doctor-patient conversations in real time and drafts them into the record in structured clinical format, along with suggested lab, medication, and imaging orders, which doctors then review before finalizing.

Within months of going live, around 800 of EMC’s doctors had AI features enabled, with roughly 200 using them actively in daily practice — uptake has been strongest in the emergency department, where fast access to a patient’s history matters most. Dr. Bella Desra Andae, a medical informatics expert at EMC, has described the old pain point plainly: legacy EHR systems interrupted clinicians as they switched between screens hunting for data. “We also see this as a game changer for the experience of both our patients and our care providers,” said Jusup Halimi, EMC’s President Director, of the rollout.

The AI layer doesn’t stop at the records desk. Chest X-rays across all eight hospitals are now read with support from Harrison.ai, a clinical AI developer, in what the companies describe as Harrison.ai’s first private-sector partnership in Indonesia. The system is trained to flag up to 124 distinct radiological findings — including patterns that may indicate cancers or chronic disease — and automatically triages scans so radiologists see the most urgent cases first. “A vital extension of our diagnostic capabilities,” is how Halimi has characterized it, noting that radiologists have responded well to both its accuracy and how cleanly it fits into existing workflow. EMC has paired this with AI-based retinal screening, now available across the network, to flag chronic disease risk from a simple eye scan.

Patient-facing AI is the newest layer. A WhatsApp chatbot tied into the EMR now handles appointment requests and general enquiries, and EMC is testing a voice bot for its general hotline that, in trials, responded to callers in under three seconds. Staff themselves have proposed the next round of projects: an AI tool to help review legal documents, and — in a very Indonesia-specific use case — a system that flags missing information in claims filed under BPJS, the national health insurance scheme, before they get rejected.

None of this happened automatically. EMC’s IT Director, Wildan Djohany, who led the original EMR rollout, points to clinician buy-in as the single biggest factor in whether any of these tools actually get used — helped along, he notes, by a Ministry of Health directive requiring Indonesian hospitals to adopt electronic records in the first place. A dedicated training and onboarding team supports doctors through each new tool, transcription accuracy on ambient listening is already estimated above 85% (trickier when clinicians code-switch between Bahasa Indonesia and English), and EMC has run third-party penetration testing while working toward ISO 27001 certification as the system handles more clinical data.

What makes EMC worth watching isn’t any single tool — it’s the sequencing. Digitize the records first, earn clinician trust, then layer in AI capability one workflow at a time, with patient-facing tools arriving last once the clinical foundation is solid. For a region where many hospital systems are only now moving off paper, that’s a more replicable playbook than a single marquee AI launch.

Source: Hospital Management Asia (19 Feb 2026); Startup Researcher (17 Apr 2026); InterSystems success story.

Tony Low covers how AI gets built and deployed across Asia-Pacific for AIAPAC.ai Insights.