Concepts
A template renders dynamically: questions can be conditional (“if previous answer was ‘yes’, show…”). The total score is a weighted sum across answered questions.
Building a template
1
Open Assessments → Templates
Click New template.
2
Add sections
Sections group related questions. They render as collapsible blocks when the assessment is taken.
3
Add questions
Each question has: type (yes/no, multi-choice, free-text), weight, and an optional conditional reveal rule.
4
Define the scoring
Map each answer to a numeric value. The template score is sum of (answer value × question weight).
5
Publish
Once published, the template appears under Take assessment for everyone in the org.
Taking an assessment
The take-assessment drawer renders the template progressively:- Conditional questions appear only after their trigger answer is selected
- Progress bar at the top
- Save-as-draft button to leave and resume later
- Submit button only enabled when all required questions are answered
Use cases
Pre-deployment review
Run an “AI release readiness” template against each app before promoting to production.
Quarterly risk review
A “risk review” template per app, scheduled by the GRC team.
Vendor model intake
A “third-party model assessment” template that gates onboarding of new providers.
Compliance attestation
Framework-specific templates that produce evidence-ready answers (EU AI Act, NIST RMF).
API
Related resources
Risk Dashboard
Assessment scores are one input to the risk score.
Evidence
Submitted assessments can be promoted to evidence.