Prediction detail
Elevated fraud, waste & abuse risk
Weighted inputs crossed the advisory band; rationale and drivers align to your published profile.
Profile fwa_prediction_profile · Model profile-1.4
Calibrated risk
Confidence & defensibility
Calibrated confidence
Level from profile rules
Weighted inputs crossed the advisory band; rationale and drivers align to your published profile.
Confidence
MEDIUM
Raw → calibrated
Index vs scaled score
Completeness
Feature coverage
Reason codes
Traceable drivers
Recommendations
From profile templates
Audit entries
Pipeline + governance
Evidence chain
Raw event through outcome
Completeness: 3 / 3 stages present
ingest
Inputs normalizedDomain adapter aligned source events to canonical signals.
score
Profile-driven scoreShared engine applied thresholds, calibration, and reason mapping.
recommend
Recommendation issuedPlaybook mapping produced a prioritized action with evidence linkage.
Top contributing signals
Weighted drivers
Weight 20% · normalized 70%
Weight 25% · normalized 70%
Thresholds
Risk bands crossed
Floor 65%
Reason codes
Structured explanations
Signals depart from peer or baseline expectations per active policy.
Recommended actions
Structured review of high-variance procedures with documentation checks.
Chronological audit
Immutable milestone trail
- SCORE COMPUTED
prediction: pred_demo_fwa_001
- RECOMMENDATION CHAIN
prediction: pred_demo_fwa_001
Processing trace
Core engine checkpoints
Feature evidence
adapter