Complete AI Training

AI agent for cloud architects

Cloud Migration Wave Planning Agent

A migration plan ordered by dependencies with risk and rollback per wave

Cloud Migration Wave Planning Agent: what goes in, what the agent does and what you get

What it does

Moving many applications to the cloud needs the right order, because moving an app before the systems it depends on breaks things. This agent reads the application inventory and dependency map, groups applications into migration waves so dependencies move together or in the right sequence, and rates each app's complexity and risk. It checks each wave for broken dependencies, where an app would run in the cloud while something it needs is still on-premises. If a wave has a gap, it moves apps between waves and rechecks. It then checks each wave's effort against team capacity and rebalances if needed. It drafts a wave plan with effort and a rollback per wave. You approve the plan. Edge case: an app depending on a system that cannot move is flagged for a network link or redesign decision.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Migration planning begins 2 USES A TOOL Read the application inventory and dependency map 3 DOES Rate each app's complexity, risk and effort 4 DOES Group apps into waves by dependency 5 CHECKS THE RESULT Does every wave avoid broken dependencies? If not: move apps between waves and recheck. Back tostep 4. 6 CHECKS THE RESULT Does each wave's effort fit team capacity? If not: move lower-risk apps to a later wave andrecheck. Back to step 4. 7 DOES Add a rollback and verification step per wave 8 YOU APPROVE Project lead approves the wave plan 9 RESULT Migration wave plan ready
Read the steps as a list
  1. Migration planning begins
  2. Read the application inventory and dependency map
  3. Rate each app's complexity, risk and effort
  4. Group apps into waves by dependency
  5. Does every wave avoid broken dependencies?If not: move apps between waves and recheck. Back to step 4.
  6. Does each wave's effort fit team capacity?If not: move lower-risk apps to a later wave and recheck. Back to step 4.
  7. Add a rollback and verification step per wave
  8. Project lead approves the wave planThe agent waits here for your OK.
  9. Migration wave plan ready

How it decides

It orders waves so no app goes live without its dependencies available, and balances waves by risk and effort.

  • Keep dependent apps in the same or earlier wave
  • Balance waves by risk and effort
  • Flag immovable dependencies for a separate decision

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Wave size and cadence
  • Risk and effort scoring
  • Business blackout dates
  • Rollback requirements per wave

What keeps you in control

It always asks you first

  • Approving the migration plan

Hard limits

  • Does not migrate anything itself
  • Flags unknown dependencies rather than guessing

It stops when

  • Done: waves are dependency-safe and approved
  • Stop: the dependency map is incomplete

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensIn planning at Ashgrove Insurance, wave two held an ordering app while its payment service sat in wave three, so the dependency check failed. The agent moved the payment service to wave two. That pushed wave two to 420 hours against 300 of capacity, so it moved two reporting apps to wave four. A mainframe-bound app was flagged for a link decision. The project lead approved the plan.

More agents for cloud architects