Turning a Knowledge Base Into a Living System
How I helped build the coaching and AI tagging behind Atlassian's customer support knowledge hub
The situation
Customer support knowledge was scattered the way it usually is, articles written inconsistently, quality varying by author, and no real system for keeping external-facing content accurate and findable over time.
Knowledge existed, but nothing was actively coaching people to write it well or keeping it organized as it grew.
What I did
Coaching knowledge into the culture, not just the content
Rather than treat this as a one-time documentation cleanup, I helped introduce KCS (Knowledge-Centered Service) coaches into the support organization, people whose job was to actively coach others on how to write and maintain good knowledge base articles, not just review them after the fact.
Building the review layer for external content
We stood up a dedicated team specifically to review knowledge base articles headed for external, customer-facing use, adding a real quality gate before anything reached customers.
Using AI to keep the system organized at scale
As the knowledge base grew, we brought in AI tagging to keep articles categorized and findable, solving the problem before it became unmanageable rather than after.
The result
A knowledge base with an actual system behind it: coaches actively developing better writers, a dedicated review layer for anything customer-facing, and AI-assisted tagging keeping it organized as it scaled.
I also created a short teaser video to help build internal excitement around the hub as it rolled out.
Internal teaser video introducing the knowledge hub.