Before the encounter: record ingestion
Patients arrive with handwritten slips from local clinics, lab reports shared as WhatsApp PDFs, and imaging histories in paper jackets. The workflow is designed to parse that material on the clinician's device and assemble a short pre-encounter brief — history, trends, and items needing attention, such as a documented condition that makes a common drug unsafe — before the consultation begins. In the public demo, this stage is shown with simulated documents.
During the encounter: a draft, in the patient's own words
In real outpatient practice, patients mix languages mid-sentence — Kannada, Hindi, and English in a single complaint. The workflow turns the conversation into a structured draft in standard clinical English: chief complaint, history, examination, assessment, and plan. The draft is a starting point for the clinician, not a finished record.
Proposed decision support — under evaluation
Alongside the draft, HealMDCare is developing decision-support capabilities: diagnostic suggestions, red-flag screening prompts phrased in the patient's language, medication-safety cross-references against the patient's documented history, and coding recommendations traced to the note. These are proposed capabilities being evaluated with clinical partners. They are not validated clinical outcomes, they make no decision on their own, and every suggestion is presented for the clinician to accept, edit, or dismiss.
Review and signature: the clinician is responsible
Every note remains a draft — visibly marked, with no auto-finalization — until the treating clinician reviews it, edits it in place, and signs it. The clinician is responsible for the accuracy of the signed record. The audit trail logs each review, edit, and signature with a timestamp.
After signing: coding recommendations
Medical coders receive ICD-10 coding recommendations in which each suggested code is anchored to the sentence in the signed note that supports it — decision support for a human coder to confirm, not automatic coding. Procedure coding is shown in the demo in CPT format for illustration; coding frameworks are adapted to local payer requirements during an evaluation.
Administration: operational view, separated by role
Role-based access is designed to give administrators throughput, review turnaround, and queue volumes across branches — while patient text, names, transcripts, and audio are outside the administrator's view.
Interoperability: designed for, scoped per partner
The platform is designed for interoperability using FHIR, HL7 v2, and C-CDA. Partner-specific EMR integrations are scoped and tested during an evaluation; we do not claim live integrations that have not been tested with the partner's environment.
Scope of the current demonstration
The current demonstration is oriented toward Indian outpatient workflows, including code-switched encounters. Support for specific languages, payers, and EMR environments is confirmed for each evaluation rather than assumed. Capabilities for other markets are future work, separate from the current India-focused demonstration.