Atlassian Team ’26 · 6 May 2026
Agentic AI in Jira Service Management
Most agentic AI today is impressive in a demo and forgettable in a workflow. The bottleneck isn’t the model — it’s the context the model has access to. An agent that doesn’t know who owns what, what’s already been decided, or what’s currently in flight can only ever be a clever chatbot. Atlassian’s Teamwork Graph — the persistent layer of people, projects, decisions, and work flowing through the enterprise — is what turns an agent from a responder into a credible colleague. This session uses Jira Service Management as the proving ground for that shift, anchored in 24 Hour Fitness’s adoption of agentic workflows.
Watch the session
Watch the full Team ’26 session (requires Team ’26 Digital pass)
Key takeaways
- Context is the moat for enterprise AI. Frontier-model quality is converging fast; what an agent knows about your business is what doesn’t.
- The Teamwork Graph compounds with every Atlassian product. Every Jira issue, Confluence page, and decision flowing through the platform makes every agent built on it incrementally smarter.
- Agents stop being a feature and start being default behaviour. The shift isn’t “click an AI button” — it’s investigation, triage, and routing happening before anyone opens a tab.
- JSM is the first proof, not the destination. The same pattern — Teamwork-Graph-grounded agents acting on real signals — generalises across every workflow Atlassian touches.
From the room
Co-presented with Rick Westbrock from 24 Hour Fitness, whose team is running this end-to-end with Dynatrace today.
Session details on events.atlassian.com →
Want me to speak at your event?
Conferences, podcasts, panels, or internal sessions — email me with the audience, dates, and format.
[email protected]