FAQ

Structure

For prospects and beta users

1. What is collective memory? How is this different from a Claude.md file?
2. Why not just use RAG?
3. How does this prevent AI hallucinations?
4. What makes aswritten different from Notion AI / Glean / other knowledge tools?
5. How do I bootstrap? What should I do first?
6. What makes a good memory?
7. How do I know it’s working?
8. Should I worry about PRs and GitHub?
9. Do I need a GitHub account? (GitHub abstraction for SMB)
10. How does the AI “think like my org”?

Pricing

1. What’s the difference between Team and Professional?
2. Can I use this on-prem? (HIPAA, GDPR, compliance)

Technical / architecture

1. How is git-native different from model-held context?
2. What is “attention dilution vs. steering”?
3. What is the extraction pipeline? How do memories become knowledge?
4. How does the ontology work? Can I customize it?
5. How does review mode work? What are lenses?
6. What does “underneath the text” mean?

---