Skip to main content
The Process

Four stages: borrower consent to an auditable credit result

Finarkein applies India's Account Aggregator consent model as specified. Data is not accessed without a valid consent artifact, and every answer carries its full reasoning trace.

Step 2

Analyst enters a plain-language query

No dashboards, pipelines, or SQL. The analyst enters a credit question in plain language, adds the borrower's consent reference, and submits it.

Finarkein checks the consent artifact, verifies that the data window fits the approved scope, and retrieves transaction data from AA infrastructure in real time.

  • Income limit checks
  • EMI-to-income ratio calculation
  • Bounce frequency and severity
  • Balance trend over time
  • Cash flow pattern analysis
Step 3

Structured answer with complete reasoning trace

Finarkein returns the answer and the reasoning behind it. The analyst can see which transactions counted as salary credits, how outlier months were treated, and precisely how the average was calculated.

This is not a score. It is a structured finding linking every answer element to the specific transaction data the borrower consented to share.

Step 4

A complete audit record for every query

Each query execution creates an immutable audit record: who asked, when, which borrower, the consent artifact, and the complete reasoning trace behind the answer.

Raw transaction data is not retained. Only the structured answer and reasoning trace enter the audit log. This meets the AA framework's data minimisation requirements and RBI guidelines.

  • Immutable audit log per query
  • Consent artifact reference on every record
  • Hash of reasoning trace verifies integrity
  • Raw transaction data not persistently stored

View four steps on a live borrower profile

Book a session to run real credit questions through the consent-to-answer flow.