CRAIM has an actual knowledge module, not just a static FAQ page.
What the knowledge base supports
The backend already supports operations for:
- creating documents
- uploading files
- searching documents
- categorizing content
- tagging
- document versioning
- workflow history
- document relations
What belongs in the knowledge base
Use the knowledge base for team-readable source material such as:
- sales playbooks
- approved positioning
- pricing and offer explanations
- objection handling
- onboarding references
- process documentation
What does not belong there alone
Do not treat the knowledge base as the only AI memory surface. CRAIM also has a separate AI workspace and observation model for runtime intelligence.
Document lifecycle
Draft
Capture initial content and structure.
Review
Let a responsible human validate accuracy and language.
Approved use
Use approved documents as reliable business evidence for both humans and AI.
Versioning and workflow
Version history matters when:
- offers change
- pricing changes
- messaging rules change
- legal or compliance language is updated
Workflow history matters when:
- the team needs traceability
- multiple reviewers touch the same document
- AI-generated updates should not silently overwrite business policy
Best practice
Keep structured facts in the right modules and use knowledge documents for explanation, policy, and operational guidance.
Examples:
- pricing tables belong in catalog or controlled operational data
- qualification criteria can live in documents, but final lead state should still be stored structurally
- message history belongs in inbox and CRM, not only in reference docs