Credit Decisioning Complete Guide
Rules engines vs decision trees, pre-bureau vs post-bureau, governance, and auditability.
See how a real application is evaluated, rule by rule, against the policies that govern responsible lending.
A multivariate rules engine: 100 rules evaluated in parallel, decision execution in under three seconds, every input and outcome captured for audit.
Engine architecture
Multivariate
Rule evaluation
Parallel · independent
Scorecard count
Unlimited concurrent
Decision engine execution time (p99)
2.4 s — Champion/Challenger
Live
Audit trail
100% inputs + outputs
99% — Decisions executed in <3 seconds
Real-time
Policy evaluation
Unlimited
Rule complexity
100% — Audit trail coverage
A decision, in flight
What you'd see if you opened the decisioning console for application APP-7421.
APP-7421 · Personal loan · $35,000
elapsed 2.41s · APPROVED
POLICY-001
Minimum age >= 18
✓ 0.2ms
POLICY-014
Income > $35,000 PA verified
✓ 0.4ms
POLICY-022
No defaults in last 5 years
✓ 0.3ms
POLICY-031
Serviceability buffer +3%
✓ 0.5ms
POLICY-047
LVR within product cap
✓ 0.2ms
POLICY-059
PEP / sanctions clean
✓ 0.7ms
SCORE-A
Credit score (Equifax)
764 (A2)
✓ 0.6ms
SCORE-B
Affordability index
0.86 — ✓ 0.4ms
PRICE
Risk-based pricing tier
Tier 2 · 6.49%
✓ 0.3ms
Outcome
Approved
auto-decisioned · no manual review
Amount
$35,000
Rate
6.49% p.a.
Term
60 months
Repayment
$686 / mo
Risk tier
Tier 2
Why architecture matters
A tree-based engine collapses under complexity. A multivariate engine scales linearly.
Tree-based
10 conditions → 1,024 branches
Exponential complexity · changes ripple through branches · duplicated logic.
Multivariate
100 rules → 100 evaluations
Linear scalability · independent rules · reusable across products.
Data sources that feed the engine
Credit Bureaus
Equifax · Illion · Experian
Identity Verification
GreenID · FrankieOne · IDMatrix
Open Banking / CDR
Basiq · Illion
Property Valuation
CoreLogic · Valocity · PropTrack
Income Verification
Payroll · accountant portals
Fraud Detection
Device fingerprint · velocity
Go deeper in this series
In-depth guides in the credit decisioning cluster.
Rules engine vs decision tree
How each encodes credit policy, where each shines, and why modern lenders lead with a rules engine.
Run it for real
Run your hardest decision through our engine
See real applications evaluated against your policies, in parallel, in seconds, with every input captured.
Learn about rules engineNext step
Want help applying this to your products and process? We’ll walk you through a tailored demo.