TadweenMed listens to the consultation — Saudi dialects, English, or code-switching mid-sentence — and turns it into a bilingual SOAP note, evidence-backed ICD-10 codes, and a clean NPHIES claim. The doctor's only job: review and sign.
In live proof-of-concept with physicians across Magrabi Health and the Ministry of Health
✓ PDPL-compliant · In-Kingdom deploymentFor every hour of face-to-face patient care, doctors spend nearly two hours on documentation and admin. The result: burnout, slower clinics, and notes written at midnight. And the global AI scribes that solved this elsewhere can't operate here — for three reasons.
Consultations mix Saudi dialects and English mid-sentence. Off-the-shelf speech AI mistranscribes and misses the clinical meaning entirely.
PDPL and hospital policy block sending recordings to foreign clouds — a hard wall for every global scribe on the market.
Manual ICD-10 / CPT coding and complex NPHIES claims cause missed charges, rejections, and slow reimbursement.
Press play and watch a real bilingual consultation become a signed, coded, NPHIES-ready record — exactly as it happens in the product today.
E11.65Type 2 diabetes with hyperglycemia✓ ValidatedI10Essential (primary) hypertension✓ Validated99214Established patient visit, moderate complexity✓ ValidatedThis is a scripted replay of a real product flow — the live pipeline runs today across mobile, clinician dashboard, coder workstation, and admin portal.
No typing. No dictation. No after-hours notes.
Ambient capture in the room — Saudi dialects, English, or code-switching mid-sentence, with automatic doctor / patient labeling.
A bilingual SOAP note in seconds, drafted by our fine-tuned clinical LLM and linked line-by-line to the transcript.
ICD-10 / CPT suggested with evidence from the conversation, validated against NPHIES rules before they ever reach billing.
The doctor reviews, signs, and pushes to the EHR — Seha FHIR, Epic, Cerner. The claim leaves clean the first time.
The doctor's only job: review and sign.
Global scribes rent their models. We train our own — on the Arabic clinical speech no one else has.
Purpose-built for the exam room, not the call center.
Fine-tuned on bilingual clinical documentation.
Every visit processed → more Saudi medical speech and clinical text → better models → a data moat that compounds.
Documentation is where revenue starts. TadweenMed makes sure the care you already deliver is the care you get paid for.
Every suggested ICD-10 / CPT code links back to the exact moment in the transcript that substantiates it.
Payer rules are checked at the point of care — so claims leave clean instead of coming back rejected.
Your billing team reviews, adjusts, and approves in one queue — no re-keying between systems.
E11.65 T2DM with hyperglycemia✓ SubstantiatedI10 Essential hypertension✓ Substantiated99214 E/M — moderate complexity✓ Rules passBuilt for this market from the first line of code — not localized after the fact.
Trust isn't a feature — it's the architecture.
PDPL-compliant processing on sovereign infrastructure. No recording ever crosses the border.
Audio, transcripts, and notes are encrypted in transit and at rest, with role-based access across every surface.
Nothing reaches the record without physician review and signature. AI drafts; doctors decide.
Every note links to its transcript evidence, every code to its justification — a full trail for compliance teams.
Pushes signed notes directly to
Seha · FHIREpicCernerJoin the physicians already documenting real Arabic and English visits with TadweenMed — and be first in line as we roll out across the Kingdom.
Riyadh, Saudi Arabia · feras@ebra.ai