Stable internet is assumed.
Outreach clinics, rural care, district services, and mobile workflows often have intermittent connectivity precisely when documentation and clinical continuity matter most.
PulseEdge · Investor overview
Clinical software assumes ideal conditions. PulseEdge is built for real ones.
PulseEdge is an offline-first, AI-assisted clinical workflow system for healthcare environments where connectivity is unreliable, governance is strict, and replacing the existing HIS is not realistic. It is designed to extend clinical operations safely, with the HIS remaining the system of record.
It is not just another AI scribe and not merely an EMR feature. It is a workflow layer built for the parts of healthcare delivery where the usual software assumptions fail.
That matters for outreach clinics, rural care, hospital extension workflows, regulated services, and public-sector environments that need modernisation without institutional disruption.
Pilot context
Bintulu
Engineering progress
7/10
Sync proof
<5s
Operating stage
TRL 3–4
The problem
The gap is not a lack of software. It is software built on the wrong assumptions.
Most clinical systems assume stable internet, central infrastructure, clean workflows, and permissive cloud AI. Many real healthcare environments do not offer any of those conditions reliably.
Generic cloud AI does not fit many settings.
Regulated clinical environments need clearer boundaries around where data moves, what AI can see, and how provenance is preserved.
Existing HIS is essential, but often core-bound.
Mature hospital systems remain indispensable at the institutional core, yet many are not designed for edge workflows, low-connectivity care, or embedded AI assistance.
Workarounds become part of care.
Paper notes, delayed entry, repeated capture, fragmented patient context, and non-governed side channels fill the gap when the software does not match reality.
This is why the opportunity is infrastructure-level rather than cosmetic. The operating assumptions are wrong, not just the interface.
Why now
AI demand is rising before deployable infrastructure has caught up.
Healthcare organisations increasingly want AI assistance, but the current market is dominated by cloud-dependent tools aimed mainly at documentation. That leaves a large gap between interest in AI and the realities of governed deployment.
At the same time, hospitals and public systems are under pressure to modernise without destabilising their core HIS environments. Replacement remains risky, expensive, and operationally disruptive.
The result is a timing window for products that can add intelligence and workflow utility without requiring ideal infrastructure or institutional reset.
The timing matters because the demand for AI is ahead of the systems that can deploy it safely in real clinical environments.
AI has become a live buying conversation.
Hospitals, ministries, and health innovation stakeholders are actively looking for credible ways to add AI to frontline workflows.
Most tools stop at the note.
Cloud documentation helpers do not solve offline operation, governed sync, continuity of identity, or real workflow structure.
Core-system replacement is usually the wrong answer.
Institutional buyers still need continuity, not just innovation. Safe modernisation matters more than a clean-sheet platform story.
Governed AI posture is moving up the stack.
Questions of visibility, deployment boundaries, privacy, and local control are becoming strategic, especially in public and regulated environments.
The thesis
PulseEdge starts from a different assumption: clinical systems must function correctly even when connectivity, governance, and workflow conditions are imperfect.
That assumption changes the product architecture, the AI posture, the audit model, and the deployment strategy. PulseEdge is designed to be offline-capable, AI-assisted, traceable, and institutionally compatible from the start.
Offline-first
The workflow is designed to continue when the network does not.
AI-assisted
AI sits inside the clinical workflow, not beside it as a generic chatbot.
Traceable
Original source, AI output, edits, and sync behaviour can be reviewed and governed.
HIS-compatible
PulseEdge extends existing institutional systems instead of forcing a risky replacement.
What PulseEdge is
A clinical workflow system, not an AI add-on.
PulseEdge is built around the visit and the surrounding clinical workflow. It supports registration, consultation, documentation, document handling, decision support, and follow-up in environments where connectivity is uneven and governance requirements are high.
It is designed to work across hospital extension workflows, clinics, and outreach contexts, while acknowledging that the existing HIS still matters and usually remains the official system of record.
In investor-readable terms, this is a workflow infrastructure play with embedded AI, not a narrow note-generation tool.
Operating model
Point of care
Outreach, clinic, bedside mobility, and other real-world care settings.
PulseEdge
System of engagement around the visit, documentation, AI assistance, and governed continuity.
Existing HIS
System of record for the institution, reporting, and historical continuity.
Visit-centred
Closer to how care is actually delivered than billing-led or document-only abstractions.
Institutionally compatible
Built to sit beside existing systems rather than require a disruptive all-at-once replacement.
Why it is different
Harder to reduce to a generic AI note layer, because the differentiation lives in the workflow and governance model.
The six existing product capabilities still matter. On an investor page, they matter because they translate into strategic defensibility and operational fit.
Deployable governance-aware AI
PulseEdge ties AI scope, visibility, and governance posture to the environment itself. That is materially different from generic permissioning layered onto a note tool.
Locked / Restricted / Standard tiersOperational continuity at point of care
Deterministic encounter identity is created at the point of care, even offline. That means the workflow can keep moving without temporary placeholders and later reconciliation risk.
Permanent PE-YYYY-NNNNN encounter identityMedico-legal defensibility
Append-only source capture plus editable working copy preserves what was originally captured even when the draft evolves. That is important for trust, governance, and institutional acceptance.
Source preserved · edits reviewableClinically safer intelligence
Flags evaluate values against patient context instead of generic thresholds alone. This makes the system more clinically credible than a document-first AI layer.
Conditions · pathway · trend · medicationsVerifiable data transformation
OCR, AI extraction, indexing, and cross-record linkage are intentionally separated into visible stages. That makes the transformation chain more traceable and more governable.
OCR → AI → Index → LinksSafer, more trusted adoption
Draft Builder surfaces missing required fields instead of masking uncertainty. That improves clinician trust and reduces the institutional risk of overconfident AI output.
Recover Missing Details · signing gated on completenessThese are not six isolated features. Together they form a governed workflow system that is harder to replicate than a generic AI documentation layer.
Evidence and traction
Early, but serious. Real engineering and institutional context are already visible.
PulseEdge is still early. This is not a scale story yet, and it should not be presented as one. But there are already meaningful signals that the product is being built against real operating constraints rather than imagined ones.
The Bintulu pilot context and the TakeCare HIS relationship ground the work in live institutional reality. Engineering progress is measurable, device-to-device sync has been demonstrated on LAN, and current readiness is being described conservatively at TRL 3–4.
- The pilot pathway suggests institutional relevance rather than abstract product exploration.
- The milestone and sync signals suggest actual execution, not just concept presentation.
- The stated TRL posture is appropriately restrained, which improves credibility.
Institutional signal
Bintulu pilot
Hospital Bintulu context provides a concrete path for real-world evaluation rather than a purely hypothetical deployment target.
System context
TakeCare HIS
The work is grounded in an operating HIS environment, which matters because PulseEdge is built to extend existing hospital systems rather than replace them.
Engineering progress
7/10 milestones
Product development is already structured and measurable, with milestone completion providing a more credible signal than generic roadmap language.
Technical validation
<5s LAN sync
Fast device-to-device sync is directly relevant to the core offline-first thesis and helps validate the system architecture under local network conditions.
Why this can scale
The architecture has the shape of a repeatable platform, not just a single-site project.
The same product logic can extend across outreach, clinic, and hospital workflows. That matters for company-building because the underlying problem repeats across institutions and markets.
One operating model, multiple care settings
The same offline-first, governed workflow architecture can support outreach, fixed clinics, hospital extension workflows, and other mixed-connectivity environments.
Southeast Asia is not the only fit
The underlying constraints — inconsistent infrastructure, strict governance, and HIS continuity — recur across Southeast Asia and other real-world health systems.
Governed AI increases institutional relevance
As sovereign and governed AI questions become more important, a workflow system that already links visibility, deployment, and AI boundaries becomes more strategically useful.
Fits both deployment and partnership models
The product can support direct deployments, hospital extension projects, and institutional or strategic partnerships where integration posture matters as much as software itself.
The scale argument is not that every hospital wants the same software. It is that many institutions face the same deployment constraints and need the same combination of continuity, governance, and usable AI.
Why the team is credible
Built from inside hospital operations, not from outside commentary.
Patek Mega is a Malaysia-based healthcare software company founded in 2019. PulseEdge is being built from the vantage point of real hospital system work, including the TakeCare HIS context at Hospital Bintulu.
That matters because judgement in this category depends on operational proximity. The hard problems are not only technical. They are institutional, workflow-driven, and governance-sensitive.
Joshua Siow leads the company with long-running experience across hospital systems and implementation environments. Clinical-domain leadership is provided by Dr. Clarice Jing Rou Siow, bringing clinical workflow, documentation, and safety discipline shaped by clinical training and pre-doctoral research experience at Harvard Medical School and Brigham and Women’s Hospital.
Credibility here comes from having lived inside the operating constraints, not from describing healthcare from a distance.
Principal partner for TakeCare HIS at Hospital Bintulu, which grounds product design in operational healthcare systems rather than generic SaaS assumptions.
Joshua Siow brings implementation and hospital-systems experience relevant to governed healthcare software deployment.
Dr. Clarice Jing Rou Siow adds workflow, documentation, and safety rigor shaped by clinical training and research experience.
Malaysia-based, MOF registered, and PDPA-aware, which matters for public-sector credibility and serious institutional engagement.
Selected technical depth
The product thesis only matters if the system behaves differently underneath.
These are selected technical design choices carried forward from the capabilities material. Here they function as validation of the thesis, not as the primary pitch.
Governance-aware AI is enforced in product state
Locked, Restricted, and Standard environments change AI scope, visibility, and governance posture together rather than leaving those concerns to policy alone.
Encounter identity is permanent from the start
PulseEdge generates deterministic, collision-resistant encounter IDs offline so continuity does not depend on later network reconciliation.
Source capture remains visible even when drafts change
Append-only source, editable working copy, and linked extraction preserve provenance across AI assistance and clinician refinement.
Flags are patient-aware, not threshold-only
Clinical flags evaluate structured values against active conditions, pathways, medications, and recent trend rather than relying only on generic ranges.
Transformation stages are visible and rerunnable
OCR, AI extraction, indexing, and relationship linking are intentionally separated so the document pipeline can be audited and re-run selectively.
Drafts surface uncertainty explicitly
Draft Builder validates against required-field schema, flags missing details, and gates finalisation on completeness instead of hiding ambiguity.
Local-first data posture
- SQLCipher local databaseEncrypted at rest with device-local protection.
- Drift ORM, schema v50Structured data model with append-only audit behaviour.
- sqlite-vec embeddingsLocal semantic retrieval path using ONNX all-MiniLM-L6-v2.
Designed for constrained deployment
- Three-tier architectureOutreach, clinic server, and hospital sync layers.
- PowerSync plus PostgresSupports fast device-to-device sync on LAN.
- WiFi relay modelmDNS discovery and ECDH session keys without requiring cloud access.
Primary path does not depend on cloud
- On-device llama.cppLocal inference remains the primary operating path.
- Groq cloud fallbackHybrid mode only, used when environment posture allows it.
- RAG over signed notesLocal embeddings remain local and are not part of sync.
Start with the right conversation: investment case, technical diligence, or institutional deployment.
PulseEdge can be reviewed as an early-stage healthcare infrastructure opportunity, as a governed technical system, or as a pilot and strategic partnership pathway.
Investor briefing · Technical diligence · Pilot and institutional partnership discussion