IT Support Agent
The agentic next generation of TracE — two agents that don’t just answer, they act: tickets, device context, incident command, and live-agent escalation.
The story, in brief.
MillerKnoll is a collective of brands running roughly 1,600 applications, where a new service-desk technician historically took about nine months to feel confident. TracE — AISE’s first production agent — cut that to eight weeks. The IT Support Agent is its agentic successor: two agents, built by an intern engineer, that go past answering questions to actually doing the work.
How it’s built.
Before there was an AISE practice, a small knowledge-retrieval copilot shipped to the global service desk in Copilot Studio, surfaced in Teams, grounded in vetted articles uploaded by hand. The point that still defines the team: it was built, maintained, and run entirely by non-developers.
Steps through 3 generations — from Era 0 to Era 2.
What AI actually does here.
Knowledge as an operating model
TracE’s real innovation wasn’t retrieval — it was treating knowledge as a living operating model: authored, senior-tech moderated, and pruned, so the base stays trustworthy as it triples in size.
Connectors turn answers into actions
Axonius, Jamf, Tanium, Jira Service Management, Outlook, Teams, OneTrust, and Genesys — each a reusable connector that lets the agents do the work, not just describe it. Device context prefilled, tickets routed, incidents run.
Teamwork Graph grounding
The technician agent grounds answers in Confluence through Atlassian’s Teamwork Graph — surfacing related articles and context by relationship, not just text match, during live incidents.
Built with the business, not for it.
The next evolution of TracE — turning the same service-desk knowledge outward to MillerKnoll associates.