PL-500 study plan

Install and configure Power Automate for desktop, then create a working desktop flow from a blank canvas

Build a flow that launches an application, reads 50 rows from an Excel file, and enters each row into a form

Add variables, loops, and conditional logic to handle blank and duplicate rows

Implement basic error handling with a block that logs failures to a CSV file instead of stopping the run

Refactor repeated actions into at least three reusable subflows with input and output parameters

Harden UI selectors so the flow still runs after the target window is resized or reopened

Add retry loops and on-block-error handling for the two most failure-prone actions

Use a scripting action to transform text and parse a value returned by the UI

Test the flow ten consecutive times and fix every intermittent failure

Call a desktop flow from a cloud flow and pass input parameters in both directions

Use the Excel and Outlook connectors to feed data into the desktop flow and email a run summary

Handle errors that cross the cloud-to-desktop boundary and surface them in the run history

Store a credential in a connection reference rather than hardcoding it in an action

Build a custom connector from an OpenAPI definition and call it from a cloud flow

Use the HTTP with Azure AD action to call a REST endpoint and parse the JSON response

Read and write records in Dataverse from a cloud flow and relate them to desktop run results

Apply a DLP policy and confirm the connector you built is classified correctly

Package flows, connectors, and environment variables into a single unmanaged solution

Replace hardcoded values with environment variables and connection references

Configure a DLP policy and verify the solution still imports under it

Set up run-history monitoring and an alert for failed cloud runs

Export the solution as managed and import it into a separate test environment

Map an as-is business process and identify the steps that are viable for automation

Write a process design document with inputs, outputs, systems, and success criteria

Define an exception taxonomy and decide retry, escalate, or terminate for each class

Design the credential and access model for every system the flow touches

Estimate build effort and flag the top three design risks

Configure a machine group and an on-premises data gateway for unattended runs

Set up a deployment pipeline that promotes the solution from test to production

Move long-running work onto queues and verify items are processed in order

Configure production monitoring and document a rollback procedure

Run a rollback drill and confirm the prior solution version is restored

Re-run missed practice items from Weeks 3 through 7 and close every gap

Build an end-to-end capstone that combines a cloud trigger, desktop flow, connector call, and Dataverse write

Complete a timed 50-question mock exam and review every incorrect answer

Document the capstone architecture and the five most likely exam traps you found