Structured clinical notes prepared by the agent, reviewed by the clinician — consistent across departments, languages, and care teams, with full audit history. 主観的情報 · 客観的情報 · 評価 · 計画.
Subjective · Objective · Assessment · Plan — the four sections, bilingual by default.
What the patient reports — symptoms, history, concerns, in their own words.
Vital signs, examination findings, and measurable clinical data.
The clinical reading — diagnosis, ICD-10 病名, and reasoning.
Treatment, prescriptions, referrals, and the next steps in care.
Subjective ／ 主観的情報. Objective ／ 客観的情報. Assessment ／ 評価. Plan ／ 計画. The same record reads cleanly to Japanese clinicians and international reviewers — one record, two languages, nothing to reconcile.
English and Japanese side by side — not an afterthought translation. The bilingual structure is the canonical form: validated at input, preserved through review, exportable downstream.
Dictation handles Kansai, Tōhoku, Kyūshū, and Hokkaido speech — clinicians dictate the way they actually speak.
Foreign patients understood in English, Portuguese, Chinese, Vietnamese, Tagalog, Nepali and more — documented in Japanese, in the same record as every other encounter.
The same SOAP record on desktop at the workstation, tablet at the bedside, and mobile in the corridor — built for how clinicians actually move between rooms.
Patient list, full SOAP sections, current medications, and interaction warnings — the clinician’s primary workstation view.
A 2×2 SOAP grid with dictation, ICD-10 search, handwriting, and one-tap Save / Review / Share. For bedside and rounds.
An AI Suggestion panel with explicit Accept / Edit / Reject — the human-review boundary made literal. Voice dictation, vitals, interaction badges.
An AI Suggestion panel drafts the SOAP narrative; the clinician sees a suggestion, not a fait accompli. Accept commits it. Edit opens it for revision. Reject discards it. Every decision is logged.
The platform never finalises a clinical artefact without that review. The AI suggestion is a tool the clinician picks up or sets down — never the record itself.
ICD-10 carries the canonical Japanese 病名 vocabulary, so every SOAP artefact exports cleanly to JAHIS-aligned systems, レセコン billing, and public-health registries — no re-keying.
A 15-minute walkthrough with KodaSōken engineering. We open SOAP, draft a clinical note end-to-end, and answer integration questions live.
AI agents prepare. Clinicians decide. Every note is audit-clean.