YIRA MedSense reads the medical histories, lab reports, hospital bills and tariffs behind every cashless and reimbursement claim in India, turns them into structured data with ICD-10 codes and drug names, checks the patient's identity and flags documents that match earlier claims. It runs in production at an IRDAI-licensed TPA today.
Medically reviewed by Dr. N. S. Santosh Varma, MBBS, FIIM, FIDM, FICCM · Medical Officer, YIRA MedSense · Updated 14 October 2026
{
"documents": ["medical_report.pdf", "invoice.pdf", "lab_report.pdf"],
"extracted": "diagnoses, ICD-10 codes, drugs, lab values, bill lines",
"identity_check": "ID photo vs hospital photo, confidence score",
"prior_record_match": "flag if fields match an earlier claim",
"output": "structured JSON to your claims system"
}Illustrative output. MedSense prepares the review; your claims system handles submission.
TPAs and insurers in India review large volumes of cashless pre-authorisations and reimbursement claims from thousands of network hospitals, each with its own formats and tariffs.
Doctor notes and medical histories still arrive handwritten and have to be read and keyed in.
Every hospital has its own room categories, packages and rates to check bills against.
Cashless requests need fast medical review while the patient is still in hospital.
Reused reports and bills are hard to spot when each claim is reviewed alone.
MedSense does not replace your claims platform or your regulator submissions. It reads the medical documents that sit behind each claim, so your team reviews structured data instead of scans. Structured output goes to your TPA portal or insurer claims system by API; MedSense prepares the review, your existing workflow and any claims-exchange submission stay as they are.
| Step | Without MedSense | With MedSense |
|---|---|---|
| Document intake | Scans and PDFs read by hand | Medical reports, lab reports, invoices and tariffs read by OCR and clinical NLP into JSON |
| Medical review | Reviewer reads every page | Diagnoses, ICD-10 codes, drug names and lab values flagged against their ranges |
| Identity | Visual check of an ID copy | ID photo compared with the hospital photo, with a confidence score |
| Fraud, waste and abuse | Found late, in audits | New documents matched against prior records; matches flagged for investigators |
| Provider pricing | Manual tariff lookup | Provider tariffs as procedure-level rate tables, contract rules applied, missing rates shown as NA |
MedSense runs in production at Health Assist Insurance TPA Pvt Ltd in India. It has parsed more than 1 lakh (100,000+) medical documents, flagged over 5,000 fraud cases for review and supported 100+ insurance operations.
Test MedSense on your own sample claim documents.
Set up document types, fields and rules for your portfolio.
Connect by API to your claims or TPA system.
Run in parallel, then switch over.
Timing depends on scope, data access and integration readiness. Client-managed deployment is available, so MedSense can run inside your own infrastructure. A field match is not a fraud determination; your team decides.
Yes. Health Assist Insurance TPA Pvt Ltd uses MedSense in production for medical document review, fraud flagging and tariff work.
Yes. It reads the same document types for both: medical histories, lab reports, discharge summaries, hospital bills and tariffs.
Yes. Clinical NLP extracts diagnoses with ICD-10 codes and drug names, including names from India's National List of Essential Medicines (NLEM).
No. MedSense reads and structures the documents and sends the data to your claims system by API. Submission through NHCX or insurer portals stays in your existing workflow.
Yes. Client-managed deployment is available, so documents can stay within your environment. Hosting and data residency are agreed per contract.
Yes. Health Assist Insurance TPA Pvt Ltd uses MedSense in production, where it has parsed more than 100,000 medical documents.
About 4–5 weeks from sandbox to go-live, depending on scope, data access and integration readiness.
Tampering and forgery checks can be built for your requirements as a scoped custom module. They are not part of the standard product today.
See MedSense on the documents and fields that matter to your claims team.