Announcing o11's backing from Y Combinator.
AI for Confluence
AI that drafts the page in the wiki.
- Draft decision docs and runbooks in place.
- Cross-link knowledge from issues, repos, and chats.
System overview
This page is the canonical reference for the Web platform service topology. Changes require approval from the Architecture council.
Architecture diagram
Architecture diagram (Lucidchart) Open in Lucid →
Services
| Service | Owner | Tier | Runbook |
|---|---|---|---|
| api-gateway | Platform team | Tier 1 | Open ↗ |
| auth-service | Platform team | Tier 1 | Open ↗ |
| billing-service | Billing | Tier 2 | Open ↗ |
| search-service | Search | Tier 2 | Open ↗ |
| reports-worker | Analytics | Tier 3 | Open ↗ |
Recent decisions
- ADR-033: Move audit log storage to S3 + Athena (decided 4/15)
- ADR-032: Replace Kafka with NATS for internal events (decided 4/01)
- ADR-031: Standardize on Auth0 for SSO (decided 3/15)
Open questions
- Confirm storage tier costs AT A. Tanir
- Document failover playbook PS P. Singh
- Update SLO targets LP L. Park
o11 for Confluence: an AI agent layer for the team wiki
o11 is an AI execution layer that runs inside Confluence. The agent drafts pages, builds decision docs, and cross-links knowledge across spaces — without sending a writer into a separate AI tool to copy-paste a draft back into the wiki an hour later.
If you have searched for AI for Confluence, a Confluence AI agent, or an alternative to Atlassian Intelligence that actually publishes pages, the difference comes down to one question: does the AI summarize the wiki or extend the wiki?
What teams use o11 for in Confluence
- Page drafting. Build runbooks, onboarding docs, and policy pages directly in the right space, with the right template and permissions from the start.
- Decision-doc creation. Generate ADRs, RFCs, and meeting notes that are properly structured for review, with the relevant Jira issues and people already linked.
- Knowledge cross-linking. Tie new pages to existing pages, Jira tickets, GitHub repos, and runbooks so the wiki stays connected instead of fragmenting.
- Page maintenance. Refresh stale pages, archive duplicates, and update labels and parent pages in batch.
Why o11 differs from Atlassian Intelligence
Atlassian Intelligence ships in Confluence as a content layer — it summarizes pages, answers questions, and suggests text inline. The execution still ends with a human creating the page, restructuring the section, or rewriting the policy. o11 is built to drive those edits. The chat is the control surface, Confluence is where the doc lands.
Built for teams that live across systems
Confluence is one part of a much larger stack — usually paired with Jira, GitHub, Slack, and a project tool like Asana or monday.com. o11 is positioned to act across all of them. A retro in Slack can become a Confluence retro page, a set of Jira follow-ups, and a linked ADR without anyone manually copying notes between tools.
Where Confluence teams typically start with o11
- Engineering teams with ADRs, runbooks, and incident reviews.
- PMs with launch plans, release notes, and roadmap pages.
- People ops with policy pages, onboarding wikis, and benefits docs.
- Customer-facing teams with playbooks and internal knowledge bases.
How o11 respects Confluence permissions
The agent operates with the user’s own account. Space permissions, page restrictions, and audit logs apply exactly the way they do for that user in the UI. Nothing is bypassed. That is what makes o11 deployable in regulated organizations that cannot trust a tool that publishes pages as an over-privileged service account.
FAQ
How is this different from Atlassian Intelligence? Atlassian Intelligence summarizes pages; o11 creates, edits, and re-structures them as part of a workflow.
Who is it for? Engineering, PM, and ops teams that treat Confluence as their source of truth.