Home › How-to guides › Automate IT and DevOps chores
How to automate IT and DevOps chores with an AI agent
I looked through 163 real cases about IT and DevOps chores and pulled out what people actually did, the prompts they used and the places it went sideways. No theory, no invented success stories.
What people actually do
- Panorama unlocks student success with Claude in Amazon Bedrock
- Model Experimentation Made Easier
- Benchling accelerates scientific discovery with Claude in Amazon Bedrock
- Verisk cuts alert volume 59% by pairing AWS DevOps Agent with New Relic MCP
- Avalara uses Now Assist and a virtual agent to cut IT tickets and provision software
- Hexaware uses Elastic AI Assistant to retrieve KPI data and charts for managed-services teams
- Thomson Reuters enhances tax guidance with Claude in Amazon Bedrock
- Deloitte Canada uses Now Assist to summarize incidents and raise IT productivity
Browse all 2754 cases to see the rest.
Prompts to start from
These are the reusable prompts from real cases. Swap in your own details, run it once on something low-stakes, and read what the agent did before you trust it.
I want an agent to investigate and propose a fix, and to wait for approval before changing production. The tool in the source case was Amazon Bedrock.
I want an agent to investigate and propose a fix, and to wait for approval before changing production. The tool in the source case was Claude.
Tools that show up most
Microsoft Security Copilot (21), Microsoft 365 Copilot (19), Claude Code (18), Custom agent (12), Claude (10), Copilot Studio (8).
Guardrails
Try it first on a staging system or with read-only credentials, and require approval for anything that changes production.
Where it went wrong, or nearly did
- Validation agent that stops after three tries: A guardrail by design: it patches failures up to three times, then leaves a blocker note for a human.
More mess: the AI agent fails page. Related: the prompt library.