Credit Decisioning Engine

Application through features, scoring, policy cutoff, and adverse-action reasons

Credit Decisioning Engine Application through features, scoring, policy cutoff, and adverse-action reasons 01 / Inputs 02 / Features 03 / Score 04 / Policy 05 / Outcome Application · applicant intake · 01 / Inputs · PII Application applicant intake PII Bureau · credit file · 01 / Inputs Bureau credit file Bank & Alt-Data · income, device · 01 / Inputs Bank & Alt-Data income, device Feature Assembly · point-in-time · 02 / Features · typed Feature Assembly point-in-time typed Scoring · scorecard / ML · 03 / Score · PD Scoring scorecard / ML PD Policy & Cutoff · hard rules · 04 / Policy · audited Policy & Cutoff hard rules audited Decision · approve / refer / decline · 05 / Outcome · logged Decision approve / refer / decline logged Adverse-Action · reason codes · 05 / Outcome · retained Adverse-Action reason codes retained applicant data PII credit file enrichment alt signals enrichment feature vector typed risk score PD cutoff verdict rules decline reasons compliance Legend primary data policy / PII async batch data store

Main Path

  • • One left-to-right path: application, features, score, policy, decision
  • • Enrichment sources fan in at feature assembly only
  • • Every stage emits an auditable, replayable record

Governed Edges

  • • Policy holds hard cutoffs separate from the model score
  • • Adverse-action reasons are derived from the same features
  • • PII enters once and is carried as typed, versioned inputs

Data-flow diagram • Built with Archify • Create yours ↗ • Hover to trace • R route • Click to focus • +/− zoom • M radar • T theme • E export