For Lenders
Underwriting 30× faster, same discipline.
One API call: composite score, signals, financial history, beneficial owners. No more six parallel registry tabs.
30×
Faster than manual research
€ 1.2 mlrd.
Underwritten volume 2025
Circle K Latvia SIA
40003521600
84of 100
AAA1.2k
AA8.4k
A24k
B98k
C32k
D11k
✓5 years of on-time filings
iCommercial pledge € 12.4M (Swedbank, 2024)
!Board change 6 months ago
02 — UNDER THE SCORE
Three layers that buy back underwriting hours.
→ 01
Composite score
Proprietary 0–100 model fusing registry discipline, financial trajectory, tax history and sector risk.
- Model versionv3.2 · 2026-Q1
- BandsAAA / AA / A / B / C / D
- Refresh cadence6h
→ 02
Signal library
147 deterministic signals: tax debt, address clusters, board changes, pledges, sanctions matches.
- F-101 → F-409147 rules
- Severityhigh / med / low / +
- Tunableby sector
→ 03
Max-loan calculator
Suggests the upper bound of responsible lending using last-12-month cash flow and sector benchmarks.
- DSCRbuilt-in
- OB connectOpen Banking PSD2
- Caution leveltunable
03 — PARTNER API
One-call dossier.
Company A makes credit decisions directly from this endpoint. You get one structured JSON with everything an underwriting session needs.
GET
/v2/companies/{rc}/dossier
Full dossier — everything
GET
/v2/companies/{rc}/score
Just score + components
GET
/v2/companies/{rc}/flags
Signals with rationale
GET
/v2/companies/{rc}/financials
5-year financials
GET
/v2/companies/{rc}/owners
Beneficial owners
POST
/v2/ob/connect
Open Banking connect
POST
/v2/kib/pull
On-demand KIB pull
GET/v2/companies/40003521600/dossier
{
"rc": "40003521600",
"name": "Circle K Latvia SIA",
"score": { "value": 84, "band": "A", "version": "v3.2" },
"flags": [
{ "code": "F-202", "severity": "low",
"l": "Board member changed 2025-11-08" }
],
"financials": { "revenue_2024": 287400000, /* … */ },
"freshness": "2026-05-08T06:00:00Z",
"sources": ["UR", "VID", "GLEIF"]
}CASE STUDY
“We cut six registry tabs and one spreadsheet. The risk committee sees exactly what the algo sees.”
Kristaps Kalniņš, Head of Underwriting · Company A
8 min
Mean decision time (was 4h)
+ 22%
Approved-application uplift
− 31%
Late-payment reduction
14d
Median integration time