Automation that can exercise judgment
Classic automation breaks the moment a step needs reading comprehension. Copilot Studio workflows fix that with three AI nodes: the prompt node (one model call), the agent node (an inline or published agent with tools and knowledge), and the M365 Copilot node (hand a step to Copilot, Researcher, or Analyst, grounded in the running user's work). Each blueprint below is a complete, buildable design.
The Morning Briefing Machine
Every workday at 7:30 AM, a workflow lists your day's meetings, has an inline agent build a brief for each one using Work IQ, and delivers the stack to your inbox before your commute ends.
Ticket Triage with Judgment
A new Dataverse ticket fires a workflow that pulls customer history, has an agent node classify the ticket against your product KB, and routes it — P1s page on-call in Teams, everything else lands assigned with a draft reply waiting for human review.
The Opportunity Research Pack
A new opportunity in your CRM triggers Researcher to build a one-page account brief — recent news, leadership, likely pain points, your history with them — attached to the record and posted to the seller's Teams before they've even noticed the lead.
Expense Review on Autopilot
Every submitted expense report gets its line items extracted by a prompt node, checked against the travel policy by an agent grounded on the policy SharePoint, and either auto-approved or routed to Approvals with the agent's per-line reasoning attached.