Signup Customization

The Signup Customization system is SCRIPE's Intelligence Engine for the self-service signup flow. Platform administrators define a question tree, each answer option carries a signal weight, and declarative RecommendationRules map answer patterns to specific editions — automatically guiding users to the plan best suited to their needs.

What Is the Intelligence Engine?

The Intelligence Engine is the recommendation system behind SCRIPE's self-service signup. Instead of presenting a static pricing table, users answer a short onboarding questionnaire. The engine matches their answers against RecommendationRules and presents a personalized edition recommendation with a localized reason. Administrators configure questions, options, and rules without code changes.

Signup Flow Overview

The full signup flow from category selection through recommendation.

OnboardingQuestion Entity

Each OnboardingQuestion represents a single step in the onboarding flow. Questions can be global (shown to all users) or scoped to an EditionCategory. Question-level branching is supported via DependsOnQuestionKey + DependsOnAnswerValue.

OnboardingAnswerOption Entity

Each OnboardingAnswerOption is a selectable answer choice for a question. Options carry scoring signals (SignalWeight) for the recommendation engine and optional visibility relevance boosts (RelevanceBoost) that re-rank options based on earlier answers.

Option-Level Visibility Conditions

OnboardingAnswerOptionCondition enables fine-grained client-side visibility control at the individual option level. Unlike question-level branching (which shows/hides entire questions), option conditions show/hide specific answer choices based on earlier answers.

SignupSessionAnswer Entity

SignupSessionAnswer persists each user's answer during an in-progress signup. It deliberately uses SignupSessionRef (a plain string) instead of a FK to the SignupSession entity to avoid cross-module coupling — the Entitlements module never imports Identity session types.

RecommendationRule Entity

RecommendationRules are the declarative matching engine. Each rule defines a ConditionJson predicate, a target edition (by tier level or specific ID), and a ScoreBonus. Rules evaluated in Priority order accumulate scores per candidate edition — the highest-scoring edition wins.

Scoring Algorithm

The recommendation engine evaluates all active rules, accumulates scores, and returns the top-scoring edition with the reason from the highest-priority matching rule.

// 1. Load all active RecommendationRules ordered by Priority ASC
// 2. For each rule, parse ConditionJson and match against SignupSessionAnswers
// 3. Matched rules add ScoreBonus to the candidate edition's total score
// 4. Sort editions by total score DESC
// 5. Return top edition with the highest-priority matching rule's ReasonEn/ReasonAr

API Endpoints

Admin endpoints for question/rule management require entitlements.manage permission. Flow and answer endpoints are public — no authentication required during signup.