TrustScore

One score from multiple data sources: identity, credit, employment, insurance, and social protection. Explainable, auditable, and available in sandbox today.

0โ€“0
Score range
0
Signal groups
0%
Thin-file support
0%
Explainability coverage

More than a credit score

TrustScore combines identity verification with credit, employment, insurance, and social data into one number your team can act on.

The scoring engine is transparent by design. Today it's built for testing and workflow design, not production lending decisions.

TrustScore - Example TrustScore
72/100
High Confidence
72

Score Breakdown

Employment stability
+14
Credit standing
+15
Insurance continuity
+11
Social protection
+8
Synthetic sandbox model
v0.1-synthetic

Five signal groups, one score

TrustScore pulls from every connected data source and traces each factor back to real evidence.

Identity

Document type, verification outcome, and confidence from the identity check.

Credit History

Payment behaviour, balances, and credit utilization from TransUnion.

Employment

Job duration, employer type, and contribution consistency from RSSB.

Insurance

Coverage duration and payment regularity from health insurance records.

Social Protection

Ubudehe classification and program participation. Especially useful for thin-file citizens.

Synthetic scoring engine

Why multi-source scoring matters

A traditional credit score is one signal from one source. TrustScore combines five data sources into one explainable number.

CapabilityTraditional Credit ScoreTrustScore
Data Sources
โœ•Single bureau score
Multi-source
Policy Use
External threshold only
Policy-ready
Thin-File Coverage
โœ•Weak or unavailable
Supported
Explainability
Limited
Traceable
Bias Monitoring
โœ•Opaque
Transparent

Transparent by design

Every score is explainable. Every factor is traceable. No black boxes.

Thin-File Ready

People without formal credit history still get a meaningful score from employment, insurance, and social data.

Sandbox-First

The engine is transparent and synthetic today so your team can validate workflows before going live.

Explainable Scores

Every factor maps back to real evidence so your team can justify decisions and review thresholds.

Multi-Source

Identity, credit, employment, insurance, and social protection all contribute when connected.

Training Roadmap

A trained model comes later, after live data quality and fairness standards are in place.

Fairness First

The current engine is explicit about its scope. Fairness tooling ships with the trained-model phase.

Score Interpretation

Your team sets the thresholds. These ranges are a starting point for sandbox testing.

020406080100
80โ€“100
Very High TrustAuto-approval for standard products
60โ€“79
Good TrustStandard approval, normal terms
40โ€“59
Moderate TrustApproval with conditions
20โ€“39
Low TrustManual review required
0โ€“19
Very Low TrustEnhanced due diligence
TrustScore

Ready to try TrustScore?

Available in sandbox today. Test scoring, thresholds, and review workflows before going live.