Power Automate

Home Tools Platforms Power Automate

A Microsoft workflow system spanning cloud actions, desktop automation, AI prompts, approvals, and governed environments.

Workflow & automation · Documented · Documentation checked 2026-10-04

Cloud flows, desktop flows, AI Builder prompts, and Power Platform environments.

Work or school account, licenses, connectors, regional AI capacity, and environment policy affect availability.

What the platform is for

Power Automate moves work through explicit processes. Cloud flows use events, buttons, or schedules; desktop flows automate machine sessions. AI Builder can add a prompt step, and approvals can hold an action for a person. Start by choosing the right execution location before deciding where AI adds value.

Know the product boundary

A work / school identity, environment, licensing, connector rights, and regional AI capacity matter. Cloud and desktop requirements differ. AI Builder’s action is Run a prompt, with older tutorials using a previous name. Generative actions are preview, not a universal capability. Copilot Studio is a separate agent-building product.

How the components connect

The trigger supplies input. Connectors perform permitted service operations. AI Builder transforms context into a draft output. Validation checks its meaning and structure. Approval actions supply a human decision. Data policies restrict connector combinations, while source permissions and retention remain separate responsibilities.

A beginner starting point

Begin with an instant sample cloud flow and invented data. Inspect its run history. Add a constrained prompt, then an approval step and rejected-case test before any send or record change. Study desktop automation only when the task actually needs a machine application.

Limits and data choices

A flow can complete with a wrong AI answer. Approvals can wait without an owner, and retries can repeat a write. Desktop UI changes can target the wrong state. This edition does not provide every AI Builder extraction model, licensing calculation, RPA deployment topology, or preview-agent walkthrough.

The technical path

Use Microsoft Learn for exact connectors, AI Builder prompts, approvals, machine setup, data policies, and tenant administration. Experienced users should design test environments, service identities, deployment ownership, and reconciliation before unattended or production use.

Feature map

Cloud flows, desktop flows, and connectors

Choose an event, button, or schedule trigger and identify where each action actually runs.

Explore Cloud flows, desktop flows, and connectors

AI Builder: Run a prompt

Map controlled inputs into an AI Builder prompt and inspect its output before taking action.

Explore AI Builder: Run a prompt

Approval flows and human decisions

Route a proposed decision to an authorized reviewer and branch on the actual approval outcome.

Explore Approval flows and human decisions

Desktop automation and machine boundaries

Understand a desktop flow’s machine, session, application, and operational requirements.

Explore Desktop automation and machine boundaries

Environments, data policies, and governance

Check which connectors may exchange information before activating an AI-enabled flow.

Explore Environments, data policies, and governance

Complete workflows

From an AI prompt to a human-approved action

Place a reviewed draft between AI interpretation and a consequential cloud-flow action.

Explore From an AI prompt to a human-approved action

A desktop draft task with a safe failure path

Create a new draft in a sample desktop application under explicit machine boundaries.

Explore A desktop draft task with a safe failure path

Coverage and availability

Models, modes & conversation

Covered: AI Builder prompts and flow types; preview generative actions are not universal.

Search, research & sources

Covered: Source retrieval depends on connectors; generated answers need evidence checks.

Files, analysis & code

Not established: Mapped records and Run a prompt; broader extraction models deferred.

Images, video, audio & voice

Covered: Media depends on connected AI services; not a native consumer studio.

Projects, instructions & memory

Covered: Business state and environment context; no general chat memory claimed.

Editing surfaces & deliverables

Covered: Cloud / desktop flow designer and inspectable run outputs.

Agents, skills, plugins & MCP

Covered: Connector actions and separate Studio agent-building boundary.

Orchestration & scheduling

Covered: Events, buttons, schedules, approvals, desktop sessions, and stop routes.

APIs, SDKs, deployment & evaluation

Separate product: Test / deployment ownership explained; every API and RPA topology not covered.

Sharing, privacy & governance

Covered: Environment data policies, connections, machine governance, and retention owners.

Plans, devices & accessibility

Not established: Licenses / regional AI capacity / devices matter; full cost and accessibility matrix not established.

Sources and verification

Documentation checked 2026-10-04. Examples are illustrative unless described otherwise.

Types of Power Automate flows · Microsoft

Run a custom prompt in a flow · Microsoft

Create and test approvals · Microsoft

Create a desktop flow · Microsoft

Desktop flow governance · Microsoft

Power Platform data policies · Microsoft

Share this page