Respocare Connect AI Launches in South Africa: A New Era of Agentic AI in Healthcare
- Matthew Hellyar
- Aug 31
- 8 min read
There is a point in the evolution of every important technology where the conversation changes.
For artificial intelligence in healthcare, that point may be arriving now.
The first chapter was about whether AI could generate text, summarise information or transcribe a clinical consultation.
The next chapter is considerably more interesting.
Can intelligent systems work across clinical information, retrieve what matters, maintain context, assist with documentation, reason across a longitudinal patient record and support a medical professional without attempting to replace the judgement of the person ultimately responsible for the patient?
On 1 September 2026, Respocare Connect AI officially launches into that next chapter.
Built in South Africa for medical professionals, Respocare Connect AI brings together a Medical AI Scribe, an agentic Clinical AI Assistant, patient information, clinical documentation and decision-support workflows inside a single clinical intelligence ecosystem.
But the most important part of the launch is not that artificial intelligence has entered another area of medicine.
AI was already coming.
The more important question is how we choose to use it.
From artificial intelligence to clinical intelligence
For the past several years, generative AI has largely been experienced through a simple interface: ask a question and receive an answer.
That interaction changed how millions of people understood artificial intelligence.
Healthcare, however, demands something considerably more disciplined.
A general AI model does not automatically understand the structure of a medical practice, the chronology of a patient record, the importance of a medication change three consultations ago or the difference between information that appears in the evidence and information that the system merely considers plausible.
That is where the emerging field of agentic AI becomes important.
An agentic system is designed to work towards a defined task by retrieving information, using tools, maintaining relevant context and determining what actions are required to complete that task.
Inside healthcare, this creates the possibility of something substantially more useful than a conversational chatbot.
It creates the possibility of a clinical assistant that can work from the patient's actual information.
That difference sits at the heart of Respocare Connect AI.
What is agentic AI in healthcare?
Agentic AI in healthcare refers to artificial intelligence systems that can perform multi-step tasks within defined clinical workflows rather than simply responding to individual prompts.
Imagine a doctor asking:
“What has changed in this patient's haemoglobin over the past six months?”
A traditional chatbot can only answer accurately if the necessary information is included in the conversation.
An agentic clinical system can instead be designed to retrieve the relevant patient records, identify the appropriate results, place them chronologically, compare the changes and generate an answer grounded in the information it has been permitted to access.
The difference is architecture.
The intelligence is connected to a system.
That distinction matters because the World Health Organization increasingly frames responsible health AI not simply as a question of model capability, but as a question of governance, safety, ethics, equity and appropriate implementation within health systems. WHO has specifically emphasised robust governance and evidence-based adoption as AI becomes more deeply integrated into healthcare.
Respocare Connect AI has been built around the same fundamental idea:
The clinician remains in the driving seat.
AI provides intelligence
The medical professional determines how that intelligence is used.
More than a medical AI scribe
Medical AI scribes are becoming one of the most immediately understandable applications of artificial intelligence in medicine.
Their value is straightforward.
A clinician conducts a consultation. The system assists with capturing the encounter and generating structured clinical documentation. Instead of spending additional time rebuilding the consultation into a medical note, the doctor reviews and approves the draft.
Respocare Connect AI includes this capability.
Its Medical AI Scribe is designed to support structured clinical documentation as part of the wider clinical workflow. The current platform combines a medical scribe, clinical AI assistant and additional patient-management and decision-support functionality rather than treating documentation as an isolated tool.
But documentation is only the beginning.
The larger opportunity appears when the intelligence can work across the patient's history.
A consultation does not exist in isolation.
Neither does medicine.
A laboratory result from February may become significant because of a medication introduced in April. A referral from another specialist may explain something that appears insignificant in today's consultation. A discharge summary, previous diagnosis, imaging report or medication history may completely change the context of the question being asked.
This is why longitudinal clinical reasoning is such an important part of the future of clinical AI.
A scribe helps document the present encounter.
A clinical intelligence system can help the doctor navigate the story around it.
The patient record should become useful again
Healthcare creates enormous quantities of information.
The difficulty is not always whether that information exists.
The difficulty is finding what matters when it matters.
Clinical notes live beside laboratory results. Referral letters sit alongside imaging. Medication histories develop over years. Different specialists see different pieces of the same patient.
The result can be an extraordinarily rich medical record that remains difficult to interpret quickly.
Respocare Connect AI was built around a simple principle:
A patient record should not merely store information. It should help the clinician understand it.
That means giving the clinical assistant access to permitted patient information through retrieval architecture rather than relying entirely on what has been typed into a chat window.
It means allowing doctors to interrogate the record conversationally.
It means helping surface chronology, changes and relevant context.
It means generating documents from information that already exists instead of repeatedly recreating the same patient story.
And crucially, it means recognising when the available information is insufficient.
In clinical AI, knowing when not to answer is part of intelligence.
Why this matters specifically in South Africa
South Africa does not need to copy the way every other country adopts artificial intelligence.
We have our own healthcare environment, our own regulatory obligations, our own infrastructure realities and our own opportunities.
The country's National Digital Health Strategy for 2019–2024 placed person-centred healthcare, integrated digital platforms and sustainable innovation among the principles required for digital transformation of the health system.
The wider national conversation around artificial intelligence is also evolving rapidly.
South Africa has been working toward a national approach to AI governance centred on responsible, human-focused deployment. The 2026 draft national AI policy was subsequently withdrawn for reworking, with Cabinet stating that the goal was to strengthen the framework and establish credible national standards for ethical AI use. The direction remains important: innovation and governance have to develop together.
Healthcare makes that requirement even more important.
Under South Africa's Protection of Personal Information Act, health information falls within the category of special personal information. Patient information therefore cannot be treated like ordinary consumer data simply because artificial intelligence makes processing it technically possible.
This is one of the central realities of building clinical AI in South Africa.
The question cannot simply be:
Can the AI do it?
We also have to ask:
Should it do it?
What information is it using?
Where did that information come from?
Can the clinician review the output?
What happens when the evidence is incomplete?
Who remains accountable for the clinical decision?
These questions are not obstacles to innovation.
They are how trustworthy innovation is built.
he human-in-the-loop principle
Respocare Connect AI is not designed to replace the doctor.
That is not a disclaimer added after the technology was built.
It is a design principle.
The system assists with information retrieval, documentation, clinical context and structured intelligence. The clinician reviews the information and retains responsibility for the medical decision.
That human-in-the-loop model becomes increasingly important as AI systems become more capable.
WHO guidance on AI for health has repeatedly emphasised human autonomy, accountability, transparency and the engagement of healthcare professionals in the design and deployment of clinical AI systems.
The objective should therefore not be an autonomous machine practising medicine.
The more compelling opportunity is an increasingly capable medical professional equipped with better access to information and better tools for using it.
That is a very different future.
And we believe it is a better one.
Intelligence as a service
There is another distinction we believe healthcare will increasingly need to understand.
Respocare Connect AI is not simply an “AI product”.
Artificial intelligence is the technology we utilise.
The service is what we build around it.
That includes clinical workflows.
Patient context.
Retrieval.
Documentation.
Governance.
Human review.
Auditability.
Support.
And the continuous work required to turn rapidly evolving AI capability into something useful inside the reality of medicine.
The value of clinical AI will not ultimately be determined by which company can place the largest model behind a chat window.
The value will be determined by how intelligently that technology is connected to the work healthcare professionals actually need to perform.
That is why we describe Respocare Connect AI as clinical intelligence for medical professionals.
What clinicians can do with Respocare Connect AI
At launch, Respocare Connect AI brings several clinical intelligence capabilities together within one platform.
Medical professionals can use AI-assisted clinical documentation, interact with an agentic clinical assistant, work with patient information and generate structured outputs from the clinical record.
The ambition is not to create another collection of disconnected healthcare applications.
It is to create one increasingly intelligent clinical workspace.
A workspace where the information captured during one part of the patient's care can become useful somewhere else.
Where a clinical note becomes part of a longitudinal record.
Where that record can subsequently support a clinical question.
Where the same underlying patient information can assist with a referral, summary, clinical report or handover without forcing the doctor to reconstruct the history repeatedly.
Every unnecessary repetition removed from that process gives something back.
Sometimes it is efficiency.
Sometimes it is clarity.
And sometimes it is time.
Buying time back for medicine
There is a tendency to describe healthcare AI almost entirely in terms of automation.
We think that misses part of the opportunity.
Medical professionals operate inside one of the most information-dense and time-constrained environments in the world.
Every hour spent documenting information that has already been documented is an hour that cannot be used somewhere else.
Every patient history reconstructed manually carries a cognitive cost.
Every fragmented record that requires another search adds friction.
If clinical intelligence can reduce some of that burden responsibly, the value is not simply that software completed a task.
The value is what becomes possible with the time returned.
More attention.
More capacity.
More time for complex patients.
More time to think.
More time to practise medicine.
Perhaps one of the most meaningful things artificial intelligence can give healthcare is not automation at all.
Perhaps it is time.
A South African contribution to the future of clinical AI
South Africa has the opportunity to participate in the development of clinical AI rather than waiting for the category to be defined elsewhere.
That matters.
Local healthcare professionals understand the realities of local medicine.
Local technology companies understand the infrastructure.
Local healthcare organisations understand the regulatory environment.
And South African clinicians should have a voice in determining what responsible AI in medicine actually looks like.
Respocare Connect AI is our contribution to that conversation.
Not because artificial intelligence should be accepted unquestioningly.
Quite the opposite.
Clinical AI should be questioned.
It should be tested.
Its limitations should be understood.
Its outputs should be reviewable.
Its claims should be challenged.
Trust in healthcare technology should never depend on excitement alone.
It should be earned.
Welcome to the new Art of Medicine
There will always be something deeply human about medicine.
The judgement formed through years of practice.
The conversation that changes because of the expression on a patient's face.
The instinct to ask one more question.
The responsibility of making a difficult clinical decision.
Artificial intelligence does not diminish those things.
Used well, it may give medical professionals more space for them.
That is the future we are interested in building.
Not artificial intelligence instead of medicine.
Intelligence in service of medicine.
Respocare Connect AI launches on 1 September 2026.
Medical professionals, practices and healthcare organisations interested in exploring the platform can request a personalised demonstration and learn more at
The technology will continue to evolve.
So will we.
But the principle at the centre of what we are building will remain the same:
The clinician leads.The intelligence assists.The patient remains the reason for both.
Welcome to the new Art of Medicine.
Respocare Connect AI gentic Intelligence for the Art of Medicine.
Comments