Healthcare that never stops.
A patient describes what is wrong, in their own words. ARIA asks the follow-up questions a nurse would, then writes it up for a doctor — and she remembers. If you said something like this six weeks ago, she brings it up before you do.
I’m a patient
Talk to ARIA. She listens, asks, and remembers what you told her last time.
I’m a doctor
Open the queue. Every case arrives already written up, with the history attached.
Every interaction, one episode.
The north-star isn’t logins or AI messages — it’s resolved care episodes that close the loop and come back as follow-up.
Intake
ARIA listens in the patient's own language and structures the story — no forms, no typing.
Review
The doctor sees the summary, vitals and deterministic red flags before the call connects.
Consult
Video when the network allows, audio when it doesn't, chat when it must — and it resumes.
Care plan
Notes, prescription and tests — filed under the doctor's registration, fully audited.
The loop closes — and care continues.
Care begins with a voice.
In a village with one bar of signal, ARIA listens in the patient's own language, turning a spoken worry into a structured clinical picture — no typing, no forms.
ARIA
Listening · Hindi
Every step of care, engineered.
From the first spoken symptom to a signed, audited care plan — a single system, not a stack of disconnected tools.
Patients speak. ARIA structures.
Symptoms described out loud, in the patient's own language, become a structured clinical picture with deterministic red flags — handed to a clinician, never acting alone.
- Hindi & regional languages
- Works on low bandwidth
- Red-flag escalation
Listening · Hindi
“Saans phoolti hai, do din se… seene mein dabav…”
A live triage queue, not a waiting list.
Patients sorted by urgency and wait time, updating in realtime. Emergencies rise to the top; the clinician sees the whole picture before the call connects.
- Realtime queue updates
- ARIA handover in context
- One-click consult start
Good morning, Dr. Rao
2 waiting · 1 urgent
2
Waiting
1
Urgent
6m
Avg wait
Rahul Yadav
Breathlessness · chest pressure
ARIA handover
Consults that survive weak networks.
Video when the network allows, audio when it doesn't, chat when it must. Consultations degrade gracefully and resume exactly where they dropped.
- Graceful audio downgrade
- Resumable sessions
- Assisted mode for ASHA workers
The database enforces access.
Row-Level Security on every table. A doctor sees only patients who've granted consent — the server decides, not the client — and every read is logged, immutably.
- Consent-gated records
- Immutable audit trail
- Encrypted document locker
Row-Level Security
enforced by Postgres, not the UI
The client requests.
The server decides.
Real health data raises the engineering standard. Identity, consent and audit sit between every client and every record — enforced by the database, not hoped for in the UI.
Every read logged
Who, what and why — an immutable trail.
Consent-gated
The patient grants; the server enforces.
RLS on every table
The public API returns nothing.
DPDP + ABDM
Consent, ABHA and FHIR shaped in.
Healthcare that never stops.
The Nirog care world
Nirog is a continuous care platform for rural India, pairing a voice-first patient app with a responsive doctor workspace.
Intake
ARIA captures symptoms by voice in the patient’s language.
Triage
A structured AI handover with red flags reaches the doctor’s priority queue.
Consult
Audio-first teleconsultation that degrades gracefully and resumes.
Care plan
Notes, prescriptions and follow-up are filed and returned to the patient.
Continuity
One consent-gated, fully audited longitudinal record — a resolved care episode.