You were handed numbers.Nobody told you what they meant.
Photograph the lab report you already have, and within minutes every result sits next to its reference range, its date and the lab that ran it, in Arabic or English.
When there is not enough to go on, you are told so rather than given a confident number.
30-day free trial. Create an account, upload a lab report and see your markers.
Health bearing
Low confidenceCoverage
based on 18 of 42 markers
Heading: improving. Three markers moved since the last panel.
In motion
Your health, in one place.
The shift
You did the test. Then nothing changed.
A PDF in your inbox is not care. What changes anything is a plan that is rebuilt every time your results move, and that is still there when you change clinic.
Honest scoring
Missing data lowers confidence. It never raises the score.
The same person, read three ways, depending on how much we actually measured.
Sufficient data
Low confidenceCoverage
based on 18 of 42 markers
Enough coverage to report a bearing, with the band shown.
Thin data
Insufficient dataCoverage
based on 7 of 42 markers
Same reading, wider band. Missing data widens doubt, it never raises the score.
Below threshold
SuppressedNot enough data to estimate
Coverage
based on 3 of 42 markers
Coverage is below the reporting standard, so no bearing is given at all.
Clinical governance
Quality you can check by using the product.
Not a wall of logos. Six rules the system enforces, each of which you can test.
Ranges come from guidelines
Reference ranges are sex-aware and traced to published clinical guidelines, not derived from the average of whoever happens to have uploaded a report.
Validated instruments only
Risk assessment uses published, validated instruments rather than in-house scoring invented for the product.
Approval is enforced, not promised
No action plan and no nutrition protocol reaches a patient until a licensed clinician has approved it. An unapproved plan is simply not visible to the patient.
Every version is retained
Any change in a patient's markers can be read against the plan that was in force when they moved.
Access is scoped and logged
A clinician can see a record only while an active care assignment exists, and every access is logged and readable by the patient.
Gaps are shown, not filled
Below the reporting standard, scores are withheld rather than estimated, and lab values the parser cannot confidently map are surfaced for review rather than discarded.
The care network
Attach to a clinic in seconds. Leave without losing anything.
The record is the patient's; the facility link is only a permission on top of it.
Join at reception
A join code or QR at the desk links you to the facility in seconds.
Match to a doctor
Choose by specialty, language and gender.
Transfer, on your terms
You approve the move, and decide whether your history travels with you.
Access you can revoke
Revoking a doctor's access deletes nothing you own.
Your access log
Every open of your record is logged. And you can read it.
Who opened it, when, what they looked at, and on what basis they were allowed to.
If a clinician can see your biology, you should be able to see that they saw it.
Access log
Dr. Layla Haddad, Internal Medicine
Opened biomarker panel
Today, 14:02
Treating doctor
Al-Zahra Clinic, Jumeirah
Viewed action plan
Yesterday, 09:41
Facility link you approved
Dr. Omar Nasser, Endocrinology
Reviewed lab report
12 Mar, 16:20
One-time share, since revoked
Illustrative entries.
The continuous loop
Your plan should change when you do.
Most plans are written once and quietly abandoned; this one is rebuilt whenever your data moves.
Measure
Labs, vitals, meals and check-ins land in one record instead of five.
Plan
A stable protocol tied to your markers, with a rotating weekly menu on top.
Adapt
New labs and check-ins re-tune it, and a clinician approves each change.
Verify
The next panel is read against the version that produced it.
Halal by default and Ramadan-aware, built on familiar regional foods, with grocery lists you can send on WhatsApp.
Health Twin
A living model of your body.
Every system, graded and tracked over time, in Arabic and English.
Three audiences
One record. Three ways to read it.
For patients
Stop carrying paper between clinics and stop guessing what a result means.
- Any lab report, read and placed against sex-aware ranges.
- A health twin across body systems, in SI or conventional units.
- A nutrition protocol, halal by default and Ramadan-aware.
- An access log you can read, and consent you can withdraw.
For clinics
Less time preparing for the consultation, more visibility over the patients you already have.
- Find any patient in your panel instantly.
- Biomarker review in a queue, with unmapped markers surfaced.
- Plans wait for clinician approval before the patient sees them.
- Join codes, doctor assignment and transfers in one console.
For payers
Know which members are drifting toward claims, while the intervention is still cheap.
- Risk banded across the cohort, with the unbandable share reported separately.
- Cardiovascular and diabetes risk from recorded vitals, not surveys.
- Cohort analytics and targeted interventions from the population console.
- The same records the clinician works from, so nothing needs reconciling.
Proof of method
You should be able to ask where a number came from.
The catalogue holds 289 markers, each with its unit, its reference interval and the direction that is harmful.
Guideline sources behind the ranges
ADA Standards of Care
Glucose, HbA1c, estimated average glucose
NCEP ATP III, ACC/AHA, EAS
Lipid panel and non-HDL targets
KDIGO
eGFR staging and albuminuria
AHA / CDC
hs-CRP cardiovascular risk tertiles
Battelino et al, Diabetes Care 2019
CGM time-in-range consensus metrics
IOM / NAM DRI
Vitamin D sufficiency thresholds
Where a citation exists it is stored on the marker itself and shown beside the result, not summarised on a marketing page.
Validated risk instruments
WHO/ISH EMR-B
Ten year cardiovascular risk, eastern Mediterranean grid.
FINDRISC
Type 2 diabetes risk over ten years.
Harmonised IDF / AHA
Metabolic syndrome criteria.
KDIGO
Chronic kidney disease staging.
Nothing is scored on inputs the instrument does not have. If blood pressure or cholesterol is missing, the cardiovascular estimate is withheld and the gap is named.
First thirty days
What actually happens, and how long it takes.
Timings below describe the platform steps. Laboratory turnaround depends on the facility and the panel.
- Day 0
Upload a panel, or order one
Upload an existing lab report as a PDF or photo, or book a panel through a connected facility. Duplicate reports are detected by file hash, so uploading the same result twice does not create a second record.
About 10 minutes
- Day 0 to 1
Results structured and checked
Values are extracted, matched to the catalogue and converted to your unit preference. Anything the parser could not map is listed as unmapped rather than dropped, and you confirm the values before they enter your record.
Minutes to structure, then clinician review
- Week 1
A plan a clinician has approved
The system drafts an action plan and a nutrition protocol from your markers, risk scores and preferences, including halal by default and Ramadan aware scheduling. A licensed clinician reviews and approves it before it becomes yours. Nothing clinical is issued unreviewed.
After clinician approval
- Day 30 onward
Re-test, and see what moved
The next panel is compared against the protocol version that was active when it was written. You see which markers moved, by how much, and whether adherence tracked with the change. The protocol is then re-tuned on that evidence.
Re-test interval set by your clinician
The difference
A different category, on purpose.
Typical consumer testing
This operating system
Al-Zahra AI
Typical consumer testing
One-time snapshot
This operating system
Continuous, adaptive loop
Typical consumer testing
A confident number with no error bar
This operating system
A score with its confidence band and coverage
Typical consumer testing
Generic recommendations
This operating system
Protocol tied to your own markers
Typical consumer testing
No clinician in the loop
This operating system
Every clinical change reviewed and approved
Typical consumer testing
Your record locked to one provider
This operating system
Transfer clinic, keep the record
Typical consumer testing
No way to see who opened your file
This operating system
A readable access log, open to the patient
Typical consumer testing
English-only, Western food and norms
This operating system
Arabic-first, halal by default, Ramadan-aware
The ecosystem
One patient-owned core. Every side connected.
Engineered to world-class standards, and built from the ground up for the GCC.
Global platforms connect data. This connects data, clinicians, and care teams, in Arabic and English.
Built for the region
Arabic-first, clinician-supervised, guideline-based, and shared only with your consent.
This platform supports preventive care and does not replace diagnosis or treatment by a licensed clinician.
Straight answers
The questions serious buyers ask, answered in public.
- What does it cost to start?
- Every member starts with a 30-day free trial with full access, and there is no permanent free tier. After the trial the entry plan is Essential at AED 367 a year, about US$100. We are clear about what is paid before you pay for it.
- What does it cost for a clinic or a payer?
- A partner facility pays nothing. Facilities keep 100% of their packages, biomarkers and tests, physicians' fees and dispensing, and Al-Zahra takes no percentage of any clinical revenue. The facility's patients subscribe to the software layer directly. Payer programmes are scoped with us case by case.
- Is this a medical device?
- No. Al-Zahra AI is a preventive health platform and clinical decision support. It does not diagnose, does not treat, and is not a substitute for a licensed clinician. Anything that looks like a clinical decision is made by a clinician, in the platform, under their own licence.
- Who reviews the clinical output?
- A licensed clinician on your facility's team. Action plans and nutrition protocols are drafted by the system and held until a clinician approves them. Every protocol version is stored with who approved it and when, so the chain of responsibility is auditable.
- What happens to my data, and who can see it?
- Your record is yours. Access is granted by you, to a facility or to a named doctor, and can be withdrawn at any time. Every open of your record is written to an access log you can read yourself: who, when, what and on what basis.
- What if the AI misreads my lab report?
- Extraction is treated as a draft, not a fact. Parsed values are shown against the original document for confirmation, unrecognised markers are listed rather than discarded, and a value you did not confirm does not drive a score. Corrections propagate to the scores that used them.
- Can I leave and take my records?
- Yes. Changing clinic never deletes your record. On transfer you choose whether your history travels with you, and revoking a facility or doctor's access ends their view of your data without erasing your own copy.
Find your bearing. Then change it.
Send us an enquiry and we reply within one business day, or step into the platform and start your health twin today.