Turn bank data into predictive cash intelligence
One API call. Plaid transactions in, 12-month cash flow projections out. Base, upside, and downside scenarios with confidence bands and LLM-powered explanations.
{
"transactions": [...],
"horizon_months": 12,
"scenarios": ["base", "upside", "downside"]
}
// Response
{
"projections": [
{ "month": "2026-04", "net_cash": 42800 },
...
],
"confidence": 0.87,
"explanation": "Revenue trend is..."
}
Three steps from raw transactions to actionable forecasts
Built for platforms that serve SMBs
Multi-Scenario Projections
Base, upside, and downside scenarios generated from the same transaction data. Each scenario includes confidence intervals so your users can plan for uncertainty.
LLM-Powered Explanations
Every forecast comes with a plain-English explanation of what's driving the numbers. Seasonal patterns, revenue trends, expense anomalies, all surfaced automatically.
Usage-Based Pricing
$0.20 per forecast call. No seat licenses, no minimums. Your cost scales linearly with your usage. Tiered plans for higher volumes.
Plaid-Native Integration
Pass Plaid transaction objects directly. No transformation layer needed. We handle categorization, deduplication, and time-series alignment.
Simple, transparent pricing
Pay for forecasts, not seats. Every tier includes the full API surface, scenarios, explanations, and confidence bands. Overages billed at $0.20/call.
The prediction layer fintech has been missing
Every platform aggregates financial data. Almost none of them forecast it. ForecastKit is the API that bridges that gap, turning historical transactions into forward-looking cash intelligence.