3 AI Prompts IT Admins Are Actually Using (With Real Examples)

The gap between AI that saves time and AI that wastes time usually comes down to one thing: how specific your prompt is.

Below are three AI prompts that IT admins are actively using in production today: one for documentation, one for user access audits, and the last for device compliance monitoring. Each one comes with the exact wording, the problem it replaced, and what the output looks like.

They were shared at Rippling's July 2026 webinar on AI for IT operations by Jonah Munoz (IT Operations Manager, Brightwheel), Bob Parsons (IT and Cybersecurity Specialist, Ghost Robotics), and Erin Merchant (Head of Customer Success at Risotto and Co-Chair of the Mac Admins Foundation).

The Mistake Almost Every IT Team Makes First

Jonah Munoz described the trap almost every IT admin falls into when they start using AI: treating it too conversationally.

"I was treating it a little too humanly," Jonah said. "I'd say something generic like 'help me do this audit' and then you kind of get sucked in, walking in one additional requirement at a time."

The fix isn't complicated. It's the same way you'd delegate work to a junior admin: a clear description of the task, the specific inputs you're providing, and what you expect the output to look like.

"Your prompts should be a very clear outline of what your expected output is," he said. So, without further ado, here are the three prompts and the results they produced.

Prompt 1: Automate IT Documentation Updates

Who uses it: Erin Merchant, Head of Customer Success and Implementation at Risotto; Co-Chair of the Mac Admins Foundation

The problem

IT documentation is chronically out of date and most teams have a mix of end-user-facing docs and internal runbooks, which sometimes means neither gets updated consistently. The manual process of identifying gaps and drafting updates is time-consuming enough that it just doesn't happen.

The prompt

"What are my top 10–20 common ticket categories from the last three months? Compare these to our existing employee-facing documentation and internal documentation. For categories where documentation exists, identify whether it's current and appropriate for the issue. For categories where documentation doesn't exist, draft it. Cite your sources so I can validate the outputs."

What changes

After running an initial comprehensive audit, Erin now runs this on a monthly cadence. The prompt includes tone, format, and review requirements, so the AI generates a batch of draft documentation updates that land roughly 90% of the way there.

"You're probably going to find that even if the documentation it pushes is not 100% correct, that the value of getting 90% of the way there means that the edits that are required of you at the end of that are going to be much slower lift," Erin said. "It's going to feel a lot easier to skim rather than deep read and deep generate."

For a team of one, she's been running this successfully for two months.

Key requirement

This prompt needs connected data sources, which means your ticketing system, your documentation repository, or preferably both. The richer the context you provide, the more targeted your output will be. If you're using Rippling AI, your connected systems are already available as context without manual exports.

Prompt 2: User Access and Offboarding Audit

Who uses it: Jonah Munoz, IT Operations Manager at Brightwheel

The problem

Every IT team needs to confirm that terminated employees no longer have active accounts, but pulling from multiple systems and reconciling them manually takes hours. Miss one account from a contentious departure and the consequences are serious.

The prompt

"I'm uploading our Rippling employee export and our Google Workspace user list. Identify every user who is currently active in Google Workspace but marked as terminated in Rippling. For each, include: their termination date, their last login date in Google Workspace, and whether their account status is consistent with our data retention policy (attached). Flag any exceptions or accounts that should have been archived or deleted based on our policy. Output this as a table."

What changes

Jonah went from a manual, multi-tab reconciliation process taking several hours to an automated workflow that pulls reports from Google Drive and produces GAM output for review. "What used to take several hours is now just a couple clicks of a button," Jonah said. "Some very serious 'let's make sure these exceptions are right' and then you can run the script."

What to expect from iteration

Getting this prompt to its final form took Jonah a full afternoon. He added requirements one by one like termination date, last login date, data retention cross-reference, and flagging accounts that had logged in after termination. He continued this until the output contained everything he actually needed.

"Once you've built it well once, you won't have to ever worry about it again," he said. The front-loaded investment pays off over every future audit.

Prompt 3: Device Compliance Monitoring

Who uses it: Bob Parsons, IT and Cybersecurity Specialist at Ghost Robotics

The problem

Ghost Robotics operates under multiple security frameworks for different government contracts. One non-compliant device can fail an entire audit, and with a large remote workforce, manual device-by-device checks aren't realistic because the cost of missing something is high.

The prompt

"Show me all devices that are currently out of compliance and identify which specific policy each device is failing."

What changes

Run through Rippling AI, this prompt returns a full list of non-compliant devices organized by policy, which allows Bob to investigate root causes (misconfigured policy, hardware limitations, deployment error) and remediate before an audit surfaces the issue. Rippling AI has direct access to device management data, so no export or upload step is required.

"It's been pretty close to 100% accurate on a lot of the returns," Bob said. "It's been a huge time saver, especially with a huge remote staff." Bob still runs a full manual review periodically as a guardrail check, but the AI-assisted workflow catches the vast majority of issues first.

Security first

Bob's team was initially cautious about what data to feed any AI tool. The shift came when Rippling AI provided a secure, contained environment for their own data. "It really helped shift that paradigm," he said. "It's always a constant battle between what's secure and what's compliant, but this gave us the foundation to build on."

What These Prompts Have in Common

Look at the three prompts above and you'll notice they don't ask AI to 'help with' anything. They specify inputs, define the expected output format, set the parameters, and tell the AI what to flag. That level of specificity is what separates a prompt that saves three hours from one that wastes an afternoon of steering.

It also means the first version of a prompt is rarely the final one. Jonah spent a full afternoon building his access audit prompt before it gave him everything he actually needed for a real audit. That investment felt expensive at the time, but now the same workflow runs every month without rework.

So, Where to Start

If you're deciding which use case to tackle first, start with your biggest pain point, not the easiest one. As Erin put it: "Hit the things that hurt the most first."

What made these prompts work, beyond the specificity of the writing, is that you can run them easily inside Rippling AI, where your underlying IT data is already connected. Jonah didn't have to explain what a Rippling employee export is. Bob didn't need to upload a device list. The AI already had access to the systems doing the work, which meant the prompts could skip the setup and get to the analysis. That's a different experience than pasting a CSV into a generic chatbot and hoping the column names make sense.

If you want to try these prompts yourself, Rippling's IT prompt library includes ready-to-use starting points for access audits, device compliance, and documentation built for the same workflows Jonah, Bob, and Erin described. And if you want more context on how IT teams are putting AI to work beyond these three use cases, this post covers the broader patterns we're seeing across Rippling customers.

These prompts were shared during Rippling's July 2026 webinar on AI for IT operations, featuring Erin Merchant (Risotto / Mac Admins Foundation), Jonah Munoz (Brightwheel), and Bob Parsons (Ghost Robotics).