Skip to content
GJ HealthcareAyush SetuABDM Integrator
All AI agents

Ayush Setu AI

Claims assistant

Draft claim narratives and pre-authorisation notes from what is already in the record.

Ask how claim narratives and pre-authorisation notes get drafted, what the billing team still reviews, and where the clinical detail comes from.

What it works on

  • Drafting claim narratives from the existing record
  • Pre-authorisation notes and the detail they need
  • What the billing team checks before submission

What it will not do

  • Promising that any claim will be approved
  • Insurer or scheme rules that are not in our approved material
The exact instructions this agent runs under
  • Never promise that a claim will be approved, and never state scheme or insurer rules that are not in the approved material.
  • Every draft is reviewed and submitted by a person — say so when it matters.

Try the claims assistant

Ask anything about how this assistant works. Answers come only from our approved material — anything else goes to a person.

See it work

Run the claims assistant on a sample record.

Every record below is invented for this page — no real patient, facility or claim. What you see on the right is the same structured output the agent produces inside the software, for a person to review.

An admission is booked. See what the billing desk gets as a starting draft.

IP admission — pre-authorisation request

Sample data

Admission
IP-1180 · Orthopaedics
Patient
62 / male, scheme beneficiary
Procedure planned
Total knee replacement, right
Clinical basis
Grade 4 osteoarthritis right knee, failed 9 months conservative management, X-ray shows joint space loss
Expected stay
5 days

Claims assistant — output

Press run to see what the agent produces.

Review a document

Put a document in front of the claims assistant.

Use one of ours, paste your own, or upload a file. You get back what is missing, what is inconsistent, and what a person should fix before it goes anywhere.

Nothing you send here is stored. The document is held only for the length of the review — it is never written to our database or our logs. Even so, please remove patient identifiers before uploading anything real.

The kind of request that comes back with a query and costs the desk three days.

PRE-AUTHORISATION REQUEST

Patient: Male, 66 yrs        Policy: Scheme beneficiary
Hospital: District referral hospital
Treating doctor: Dr. Suresh

Diagnosis: Knee pain
Procedure requested: Total knee replacement (right)
Estimated stay: 5 days
Estimated cost: As per package

Clinical note: Patient has knee pain since long time. Not improving. Advised surgery.
X-ray done.

Review findings

The findings will appear here.

Ready for the real thing?

The agent above answers questions. A demo shows it on your own workflows, with your masters and your speciality templates. Tell us how your facility runs and we will bring something relevant to the call.

  • You get a reference number immediately
  • A specialist reads it, not a bot
  • A call within one working day
How should we reach you? *

Fields marked * are required. You will get a reference number like WEB-2026-000145 as soon as this reaches us.