A clinical documentation and patient-record system for South African practice.
The guide 3 min 10 . press play to run
Pilot and Scale proposal . 16 September 2026 Bioeconomy and Medical Technologies
drscribble.life
SUMMITWEBCRAFT (Pty) Ltd
Good morning, and thank you for the time. I am Irvan Damon, founder of SUMMITWEBCRAFT. Dr Scribble is a clinical documentation and patient-record system for South African practice. The single most important thing I can tell you in the first ten seconds is that it is live in production today. It is not a concept, it is not a prototype, and it is not a demonstration built for this panel. You can open it yourself at drscribble.life with no login required, during this call or after it. We entered under Bioeconomy and Medical Technologies rather than ICT deliberately. This is a medical technology that changes how the clinical record is captured, coded, carried and reached in an emergency. Artificial intelligence is the enabling layer underneath it. It is not the product.
The problem
Four things South African general practice loses
1
The note
Written after hours, thinly, or not at all. The consultation is lost to the patient sitting in front of the doctor.
2
The history
Records stay at one practice. They do not follow the patient to a hospital, a specialist, or a roadside.
3
The language
Summaries given in English, to a population with eleven official languages. A plan that is not read is not followed.
4
The claim
Codes reconstructed from memory after the fact. That is where practice revenue leaks and where rejections begin.
Every one of these is a documentation problem wearing a clinical disguise.
Four losses, and they compound. First, the note. It gets written after hours, or thinly, or not at all, because the doctor is choosing between the record and the patient in front of them. Second, the history. Records sit at one practice and do not follow the patient. When someone is admitted, or travels, or collapses at a roadside, the history that would keep them safe is unreachable. Third, the language. We hand patients a plan in English in a country with eleven official languages and a wide literacy spread. A plan that is not read is a plan that is not followed. Fourth, the claim. Codes are reconstructed from memory afterwards, and that is exactly where revenue leaks and where rejections originate. Each of these looks clinical. Every one of them is a documentation failure.
The solution, as it exists today
One consultation. Six outputs.
Live in production today. Artificial intelligence is the enabling layer underneath it. It is not the product.
My Results 1 min 06 . the vault
Clinician-confirmed note
Structured, reviewed and confirmed by the treating doctor before it becomes a record.
Summary in the patient's language
Read aloud where literacy or eyesight is a barrier.
Referral letter
Populated from the consultation, ready to send.
ICD-10 codes for the claim
Bridged from a licensed terminology source, confirmed by the clinician.
My MedAlert Profile
Allergies, chronic medications, conditions, blood type. Carried by the patient.
Emergency ScribeScan
One permanent, revocable QR. No app, no login, read-only.
The two highlighted surfaces are patient-held. Audio is transcribed and immediately discarded, and nothing becomes the clinical record until the treating doctor confirms it.
One spoken consultation produces six things. A clinician-confirmed clinical note. A patient summary in the patient's own language, which can be read aloud. A referral letter. And the ICD-10 codes the claim needs. The last two are held by the patient rather than the practice. My MedAlert Profile carries allergies, chronic medications, conditions and blood type. Emergency ScribeScan is one permanent, revocable QR on a printed card or a phone, which a first responder scans with no login and no app. Two things to hold on to. The audio is discarded immediately and never stored. And nothing becomes a clinical record until the treating doctor confirms it.
Criterion 1 . Technical merit and innovation quality
The model is never asked to return a clinical code
Published evidence on ambient clinical scribes reports occasional clinically significant inaccuracies. Our architecture is the answer to that failure mode. It is not a claim that the failure mode does not exist.
01
Confirmed note
The doctor reviews and confirms before anything else happens.
02
Plain language out
The model returns clinical concepts as language. Never a code.
03
Licensed source codes it
A licensed terminology source supplies every code, bridged to ICD-10.
04
Clinician confirms
The doctor accepts the code before it is stored against the claim.
Row-level security enforced and verified in production
A cryptographic hash on every record, so later alteration is detectable
Privacy policy published and reachable without a login
All evidenceable on request.
This is the slide I would most like you to remember. The published literature on ambient clinical scribes reports occasional clinically significant inaccuracies. I am not going to stand here and tell you that cannot happen. Instead I will show you the architecture we built because it can. The model is never asked to return a clinical code. It returns language. A licensed terminology source supplies every single code, bridged through to ICD-10. And the clinician confirms before anything is stored. So the failure mode is contained by design rather than by hope. Underneath that: row-level security verified in production, a cryptographic hash on every record so any later alteration is detectable, and a published privacy policy. All of it evidenceable on request.
Criteria 2, 3 and 4 . Pricing, feasibility, integration, scale
Public pricing. Configuration, not development.
Trial
R0
Seven days. No card required.
Clinician
R750
Per user, per month
Enterprise
Custom
Onboarding, support and terms
Live and public today. No VAT charged on any plan. Nothing here is quoted specially for this panel.Check it on drscribble.life
Feasibility and integration
One to two months to pilot start. The pilot requires no new build.
Browser-based. No installation, and no hardware beyond a device the clinician already owns.
Runs beside what the practice already uses. Nothing is ripped out and nothing is migrated.
The patient side needs no app and no account.
Scale
13,000 general practitioners and 8,000 specialists in South African private practice.
Five live tenant deployments on one platform today. A new site is onboarding, not engineering.
SNOMED CT to ICD-10 coding live in production, built for South African claim practice.
Namibia, Botswana and Eswatini share ICD-10 claim practice.
A global scribe can transcribe a South African consultation. It will not build a South African claim.
Pricing is live and public. A seven day trial with no card, Clinician at seven hundred and fifty rand per user per month, and Enterprise on custom terms. No VAT on any plan. None of that was invented for this room. On feasibility, the pilot needs no new build. Deploying to a new site is configuration rather than development, which is why one to two months is comfortable. It is browser-based, it runs beside what a practice already has, and the patient side needs no app at all. On scale, the private market is around thirteen thousand GPs and eight thousand specialists, and we already run five live tenant deployments on one platform. But the defensible scale argument is the coding layer. SNOMED CT to ICD-10, live in production and built for South African claim practice. That is the layer a global competitor will not build for this market.
The pilot pathway . Part one
How we reach a rural clinic tomorrow
Click Learning places technology, connectivity and trained local people into under-serviced South African communities. We have already delivered into that network together, teaching mathematics in rural Eastern Cape schools Click reached and equipped.
The bridge 23 sec . subtitled
Communities that already trust them
Health data asks far more of a community than a maths lesson does. A stranger with a laptop does not get that trust.
Trained local facilitators
Unemployed youth already trained to support technology in these settings. In a clinic the role stays non-clinical.
Devices and connectivity, proven
Not modelled. Running at these exact sites today, which is how the schools work at all.
District relationships
The protocols and credibility that opened the school door open the clinic door. They take years to build from nothing.
The schools themselves
School health screening happens on paper and is lost. Same building, same day, a record that follows the child instead.
The geography
The clinic and the school serve the same families. A child in a Click classroom has a mother at that clinic.
None of this has to be built. It is standing, and it is already ours to work with.
This is how a rural pilot actually happens, and it is the part I would most like you to test me on. Click Learning is an organisation that puts technology, connectivity and trained local people into under-serviced schools. I have already worked inside that network, teaching mathematics in rural Eastern Cape schools that Click reached and equipped. Six things they bring. Community trust, which health data needs far more of than a maths lesson. Trained local youth facilitators, whose role in a clinic stays non-clinical. Devices and connectivity proven at those exact sites, not modelled. District relationships that took years to build. The schools themselves, where health screening currently happens on paper and disappears. And the geography, because the clinic and the school serve the same families. None of that has to be built. It is standing.
The pilot pathway . Part two
Three pilots in motion, on one platform
Pilot one . From end September
A South African medical scheme
General practitioners across the scheme's network, in the environment where claims are actually adjudicated.
Tests the coding and claim path end to end, from consult to submitted claim.
And My Results, the patient's laboratory and imaging vault, held for life and visible to the next clinician.
Working with a senior administrator. The scheme is not named at this stage.
Pilot two . With Click Learning
Rural clinics, Eastern Cape
The clinic captures. A registered practitioner confirms. The referral reaches an off-site doctor the same day.
Click brings the sites, the facilitators, the devices and the connectivity already proven there.
We bring the platform at no licence cost, the clinician training and the local-language work.
Consent and the boundary of the facilitator role agreed in writing before a single consultation is captured.
Pilot three . Via an ambassador
Mining clinics
A doctor, entrepreneur, author and philanthropist who has offered to initiate a pilot with a mining house.
The same bridge, in occupational health. A mine clinic hours from a specialist, and the specialist in the city.
Tests the pathway where the distance is greatest and the workforce is most concentrated.
Already in clinicians' handsPrivate general practitioners, physiotherapists and allied practitioners are testing the interface now, in their own time and their own practices. That is where the usability findings come from, and it is the reason the product changes weekly.
Not one figure from any of these is claimed today. Measuring them is the reason the pilots are needed.
Three pilots, and they test different things. The first begins at the end of September with a South African medical scheme, running with general practitioners across that scheme's network. That means the coding and the claim get tested where claims are actually adjudicated rather than in a simulation. It also tests My Results, the patient's laboratory and imaging vault, which holds reports from any laboratory and makes them visible to the next clinician. I am not naming the scheme at this stage. We are working with a senior administrator there. The second is the rural pathway on the previous slide, through the Click Learning network. The third has been offered by one of our ambassadors, a doctor, entrepreneur, author and philanthropist, who has offered to initiate a pilot with a mining house. That is the same bridge again, in occupational health: a mine clinic hours from a specialist, and the specialist in the city. And underneath all three, the platform is already in clinicians' hands. Private GPs, physiotherapists and allied practitioners are testing the interface in their own time. That is where the usability findings come from. Not one figure from any of this is claimed today. Measuring them is the reason I am asking for a pilot rather than a purchase order.
Criteria 5 and 6 . Team capability, and social and economic impact
Who built it, and what it changes
The team
Trading since February 2017. Recurring revenue from contract product development funded the entire build, with no outside capital.
Irvan Damon designed and built the full pipeline, holds the SNOMED CT affiliate licence and is the registered POPIA Information Officer.
Five live tenant deployments run on the same platform today.
The constraint, stated plainly: this is a single-founder company. A structured pilot is how the operational evidence a larger buyer requires gets built.
The impact
Emergency access to allergies and chronic medications for a first responder. A direct patient-safety intervention.
A clinical record that follows the patient between practices rather than fragmenting.
Treatment plans in the patient's own language, read aloud where reading is a barrier.
A referral bridge that lets a rural clinician hand a specialist something usable.
Fewer rejected claims, which for a small practice is the difference between viability and closure.
On team. SUMMITWEBCRAFT has traded since February 2017 and earns recurring revenue from contract product development. That revenue funded this entire build with no outside capital. I designed and built the pipeline, I hold the SNOMED CT affiliate licence, and I am the registered POPIA Information Officer. We run five live tenant deployments on the same platform today. And I will state the constraint rather than let you find it. This is a single-founder company. One of the reasons I want a structured pilot is to build the operational evidence base a larger buyer is going to demand. On impact, the one I would underline is the emergency profile. A first responder seeing a patient's allergies and chronic medications at a roadside is not a productivity feature. It is a patient-safety intervention.
The ask
What we are asking of The Innovation Hub
1
Pilot support
To run the eight-week design properly and measure it honestly. No indicator in the pack is claimed today.
2
Clinical validation
Independent clinical oversight of the coding accuracy measurement. That is the specific gap the pilot exists to close.
3
IP strategy assessment
Covering the South African terminology mapping and the emergency-access method.
4
Network introduction
Into the programmes and partners where a rural clinic pathway fits.
See it working.No login required.
drscribble.life
Irvan Damon . Founder, CEO and CTO
design@summitwebcraft.co.za . +27 73 644 0126
SUMMITWEBCRAFT (Pty) Ltd . Reg. 2017/374199/07
Four things. First, pilot support, so the eight-week design runs properly and is measured honestly. Every indicator in our pack is a figure to be measured during the pilot. Not one of them is claimed today, and that is the reason the pilot is needed. Second, clinical validation. Independent clinical oversight of the coding accuracy measurement, which is the one gap the pilot exists to close. Third, an IP strategy assessment covering the South African terminology mapping and the emergency-access method. Fourth, introduction into the programmes and partners where a rural clinic pathway fits. And a standing invitation. The system is live at drscribble.life with no login. Open it during the questions if you want to. Thank you.