One API. Four markets. 200+ behavioral signals.
Panthera connects to alternative data sources across Vietnam, Indonesia, the Philippines, and Japan, extracts per-market behavioral features, and returns an interpretable score with reason codes through a single REST integration.
Alternative signal sources your bureau can't touch.
Panthera connects to mobile payment networks, e-commerce platforms, utility billing systems, and payroll processors through direct partnerships and standardized data feeds. No bureau data required at origination.
Mobile Wallet Transactions
MoMo, ZaloPay (Vietnam), GoPay, OVO, DANA (Indonesia), GCash, Maya (Philippines), PayPay (Japan). Transaction velocity, peer transfer patterns, top-up regularity.
Utility Bill Payment Records
Electricity, water, and telecom billing histories from national and regional utilities. On-time clearance rate, payment delay distribution, seasonal adjustment.
E-Commerce Purchase History
Shopee, Lazada, Tokopedia, and regional platform transaction records. Installment repayment behavior, purchase regularity, return patterns as income proxy.
Payroll and Income Transfers
Payroll deposit timing, salary regularity index, employer consistency signals, and bonus cadence from banking transaction metadata with borrower consent.
Per-market models. Not a pooled approximation.
A single model trained across markets produces worse outcomes in every market than country-specific models. Distribution shift is real. Panthera maintains separate gradient boosting models for each of its four markets, tuned to local data availability, economic seasonality, and regulatory constraints.
One integration, four market models.
The Panthera API accepts alternative data inputs and returns a score with reason codes in under 800ms P99. No bureau pull required. Lenders route requests to the appropriate market model via a single endpoint parameter.
# POST /v1/score - request body
{
"market": "indonesia",
"applicant_id": "ap_7k2m9p",
"data_sources": [
"mobile_wallet",
"utility_billing"
],
"wallet_token": "tok_gopay_...",
"utility_token": "tok_pln_..."
}
# 200 response
{
"score": 748,
"band": "B",
"reason_codes": [
"R1: High wallet velocity",
"R2: Consistent bill clearance",
"R3: Income regularity"
],
"latency_ms": 312
}
Ready to integrate alternative credit scoring?
Talk to our integration team about sandbox access, data partnership requirements, and what a pilot deployment looks like in your market.