Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for cloud architects

Virtual Machine Right-Sizing Agent

Lower cloud and virtualization cost with no loss of performance

Virtual Machine Right-Sizing Agent: what goes in, what the agent does and what you get

What it does

Machines are often oversized and cost more than they need to. This agent reads 60 days of CPU, memory and disk use for each virtual machine and finds ones that are oversized or idle. It proposes a new size for each one, then tests the proposal against peak-day usage so that month-end or batch jobs do not suffer. After a resize it rechecks performance to confirm that the saving did not cause slowdowns. The owner approves each resize or shutdown. Edge case: a machine is idle for 50 days but spikes on quarter-end, so the agent keeps it and records the reason.

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, continueApprovedYes, continueNoNo 1 STARTS WHEN Monthly usage review 2 USES A TOOL Read 60 days of CPU, memory and disk metrics 3 DOES Find idle and oversized machines 4 DOES Propose a smaller size for each 5 CHECKS THE RESULT Does the proposed size hold the peak day withheadroom? If not: pick the next larger size or keep the machine asit is. Back to step 3. 6 DOES Estimate savings and risk for each change 7 YOU APPROVE Owner approves each resize or shutdown 8 USES A TOOL Apply the approved changes in a set window 9 CHECKS THE RESULT Is performance normal one week after the change? If not: revert to the old size and update the proposal.Back to step 3. 10 RESULT Savings report
Read the steps as a list
  1. Monthly usage review
  2. Read 60 days of CPU, memory and disk metrics
  3. Find idle and oversized machines
  4. Propose a smaller size for each
  5. Does the proposed size hold the peak day with headroom?If not: pick the next larger size or keep the machine as it is. Back to step 3.
  6. Estimate savings and risk for each change
  7. Owner approves each resize or shutdownThe agent waits here for your OK.
  8. Apply the approved changes in a set window
  9. Is performance normal one week after the change?If not: revert to the old size and update the proposal. Back to step 3.
  10. Savings report

How it decides

A machine is oversized when peak use stays below a set share of capacity. A new size must still hold the peak with headroom.

  • Call a machine oversized when peak use is under 40% for 60 days
  • Require 20% headroom over the peak
  • Keep machines with known seasonal peaks
  • Revert on any sustained slowdown

Make it yours

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

  • Metrics period (default: 60 days)
  • Headroom (default: 20%)
  • Idle definition
  • Machines excluded

What keeps you in control

It always asks you first

  • Owner approves every resize or shutdown

Hard limits

  • Never resizes or deletes a machine itself
  • Never touches production without approval

It stops when

  • Done: approved changes are applied and performance is normal
  • Stop: metrics are missing for the period

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 happensA reporting server used 22% CPU at peak over 60 days on 16 cores. The agent proposed 8 cores and tested the quarter-end peak at 61%. The check passed and the saving was $310 a month. After the owner approved the resize, a week later the recheck showed normal response times. A second server failed the peak test and stayed as is.

More agents for cloud architects