FAME Score is a cash-flow-first underwriting score for thin-file Indian borrowers — built on 11 million Account Aggregator + GSTN + bank-statement profiles, calibrated against 36 months of repayment outcomes.
Every FAME Score returns the band, the bureau-equivalent, six explainable drivers and a recommended action — wired into your LOS in a single call.
Trained against actual 36-month repayment, recovery and write-off events from partner lenders — not survey panels or synthetic data.
FAME Score is most useful exactly where bureaus aren't — first-time borrowers, sole-proprietor MSMEs, new GSTINs, kirana digitisation.
Every band carries six drivers and the source rows that moved them. Your credit officer can defend any score to an auditor.
A GST default, a bounce, a new enquiry — the score updates the moment it happens, and the alert lands in your queue.
Returns a policy-aware suggestion — approve / refer / decline, with amount and tenor bounds tuned to your underwriting matrix.
Designed to layer on CIBIL / Experian / CRIF — most customers run FAME + bureau in tandem with a policy gate on either.
Four agents, one request. Average wall-clock: 4.1 seconds.
Borrower approves AA + GSTN consent in a single flow. PAN, GSTIN and mobile keyed in.
~ 90 s · borrower-side11 bank accounts, 24m of GSTR-1/3B and 3-bureau CIC pulled, deduplicated, normalised.
~ 2.1 s · serverFAME v4 ranks 412 features, scores six drivers, produces the band and a citation set.
~ 820 ms · modelPolicy engine returns approve/refer/decline + amount, tenor and rate bounds.
~ 60 ms · policyREST, webhook or LOS plugin. SDKs for Python, Node, Java.
API keys in 24 hours for licensed lenders. Sandbox available for evaluation.