Praedias

A Questionnaire That Adapts to Every User

The questionnaire needed to show only the questions relevant to each person's exact situation, with branching complex enough that hard-coding every path wasn't realistic.

Solution

Uptech built a rule engine that evaluates each question's visibility in real time against a serialized condition and the user's profile, resolving dependencies across the entire questionnaire through a dependency graph. New branching scenarios can be added as configurable content rather than code changes.

‍

Keeping the Task Plan in Sync as Answers Change

Users often revise their answers partway through, and the platform needed the resulting task plan to update instantly without disturbing work already completed.

Solution

Every answer change triggers a full re-evaluation against an in-memory projection of the user's answers, and the engine classifies each affected task as create, restore, pause, or remove. Finished work is protected, and in-progress tasks are paused rather than deleted.

‍

Preventing Data Loss During Multi-Step Editing

As sections of the questionnaire appeared and disappeared based on new answers, there was a risk of losing or corrupting data the user had already entered.

Solution

Answers move through a three-state lifecycle of staged, active, and suspended rather than being deleted. New answers are validated before promotion to active, and answers tied to a hidden branch are suspended and automatically restored the moment that branch becomes relevant again.

‍

Turning Raw Answers into a Personalized, Deduplicated Action Plan

Instead of a generic checklist, users needed a task list built from their specific situation, including repeating items like multiple properties or bank accounts, without duplicate or conflicting tasks.

Solution

The same rule engine matches a library of task templates against the user's answers, generating one task per repeating entity and deduplicating overlapping answers so re-editing details updates existing tasks instead of creating copies. A CQRS architecture keeps task generation, answer processing, and access provisioning decoupled and reliable.