Recommendation Engine

Updated February 22, 2026

Recommendation Engine describes how recommendation scoring, confidence, rationale text, and evidence metadata are generated in Unfair.

Why it matters

This term explains how the model ranks options and why some bundles are shown before others.

Unfair uses this term in onboarding signals, ranking output, rationale snippets, and recommendation refresh flows.

Deterministic Scoring describes how recommendation scoring, confidence, rationale text, and evidence metadata are generated in Unfair.

Evidence Tier describes how recommendation scoring, confidence, rationale text, and evidence metadata are generated in Unfair.

Supplement Stack describes how stacks are structured around goals, templates, activation states, and cycle timing in Unfair.