Why KnowStack? The Complete Workflow How Features Connect Use Case Examples Getting the Most Value
Creating Your Account First Login & Dashboard The General Flow: From Data to Knowledge Understanding Navigation
Website Crawling Email Integration Document Upload Meeting Transcripts Telegram Integration Slack Integration
Generating a Knowledge Base Iterative Generation
Viewing & Navigating AI Search Editing Content Version Management Exporting Access Control
Inviting Team Members Roles & Permissions
Profile Settings Company Settings Notifications
Plans & Pricing AI Usage & Billing
Data Protection Account Security Secure Integrations
Integration Overview Managing Integrations
Overview Creating a Test AI Question Generation Assigning Tests The Candidate Experience AI Grading and Manager Override Analytics Releasing Results Abandonment and Retakes FAQs Troubleshooting
Tracking Activity
Getting Help Common Issues
Common Questions

AI Question Generation

KnowStack uses the same OpenRouter-backed AI pipeline that powers generation and search elsewhere in the platform. Questions are drafted directly from your Knowledge Base content, so the answers the AI considers correct are always grounded in what you have documented.

How Grounding Works

  • The selected KB (and sections, if specified) is fed to the model as authoritative context
  • Each generated question references specific KB content so correctness can be verified
  • The AI does not use external facts; if your KB does not cover a topic, no question is created about it
  • Essay and short-answer questions include a model reference answer for downstream grading

Cost and Regeneration

Generation cost is shown up front. If you are unhappy with the first pass, you can delete questions and regenerate with a tighter section scope or a different mix of question types. AI usage for each attempt is logged and billed separately.