The cash-flow decisioning layer for your lending stack
From bank transaction data to a fully explainable credit decision — in a single API call. No score supplementation, no black box. A complete decisioning layer designed for lenders.
Three-stage decisioning pipeline
Raw transaction data enters. A structured decision with reason codes exits. The 24-month lookback window is the foundation — long enough to identify patterns, short enough to stay current.
The pipeline processes 24 months of bank transaction history, extracting velocity signals across income consistency, expense discipline, liquidity cushion, and payment reliability — then outputs a decision score and Reg B reason codes.
How each signal category works
Each signal category reflects a distinct dimension of borrower financial behavior. Expand to understand the methodology.
Income consistency measures whether a borrower receives income-categorized deposits at regular intervals over the 24-month lookback window. The model identifies recurring deposit patterns — weekly, bi-weekly, semi-monthly, or monthly — and scores both the frequency regularity and the magnitude stability.
Gig workers, freelancers, and cash earners often show irregular deposit timing but consistent aggregate monthly income. The model handles this by measuring income over rolling 30-day windows rather than requiring payroll-style periodicity. A borrower with variable gig income can score well on income consistency if their monthly inflow is stable within a reasonable standard deviation.
Why it matters: FICO requires tradeline data — a borrower must have used credit to have a score. Income consistency from cash-flow data is signal-positive regardless of prior credit use.
Expense discipline is computed as a coverage ratio: recurring fixed-obligation outflows divided by recurring income inflows, measured over the 24-month window. Fixed obligations include rent/mortgage, utilities, insurance, and subscription-category recurring charges.
A high coverage ratio (obligations consuming most income) signals stress. A stable ratio over time — even at a moderately elevated level — signals a borrower who has found a sustainable equilibrium. The trend matters as much as the level: a borrower whose ratio is improving over 24 months signals upward trajectory.
Liquidity cushion measures the average minimum balance observed in the 72 hours before the next expected income deposit — the lowest point of each pay cycle. This captures how much buffer a borrower maintains at the economically stressful point of their cycle.
Even a small persistent cushion (e.g., $50-200) is a meaningful signal: it indicates the borrower does not fully deplete each paycheck, suggesting intentional or habitual reserve behavior. NSF events are tracked as negative signal — each NSF in the 24-month window is a data point against the cushion score.
Payment reliability measures the longest uninterrupted streak of successful recurring outflows — the opposite of NSF or return events. Recurring outflows that have never been interrupted over the 24-month window receive the highest signal weight.
This captures the same behavioral dimension as a FICO payment history, but using a different underlying data source. A borrower who has paid rent, utilities, and at least two subscriptions without interruption for 24 months is demonstrating the same discipline that on-time credit card payments would demonstrate — without requiring prior credit access.
Embed in your existing LOS in minutes
No new infrastructure required. Lendiro is a REST API that sits alongside your existing loan origination workflow — not a replacement.
Three core endpoints
The Lendiro API is REST + JSON. Authentication is Bearer token. All endpoints return structured decision data with audit fields.
// POST /v1/decisions — Submit transaction data for decisioning
// Request
{
"applicant_ref": "app_8f2a3c91",
"lookback_months": 24,
"transactions": [
{
"date": "2024-05-15",
"amount": -850.00,
"category": "rent",
"recurrence": true
}
],
"requested_loan": {
"amount": 5000,
"term_months": 24
}
}
// Response — 200 OK
{
"decision_id": "dec_7e4b1f20",
"score": 74,
"recommendation": "approve",
"reason_codes": ["CF-01", "CF-04"],
"signals": {
"income_consistency": 0.82,
"expense_discipline": 0.71,
"liquidity_cushion": 0.68,
"payment_reliability": 0.91
},
"audit_log_ref": "audit_3a9c7d"
}Start with the documentation
The quickstart guide walks through authentication, your first API call, and integrating decision responses into your LOS workflow.