IT playbook · AI Employee: Ivy

Hardware Troubleshooting

60% of hardware issues resolved without IT touch

Install this playbookBack to marketplace
The problem

Every hardware issue starts as a mystery: "my laptop is slow," "WiFi keeps dropping," "external monitor won't work." The user tries turning it off and on, gets bored, files a ticket. IT asks the same 10 diagnostic questions, walks the user through the same checklist, and half the tickets resolve with "try a different cable" or "clear DNS cache" — after a two-day wait. IT loses hours on tier-one triage; users lose faith that IT can help at all.

At a glance
Trigger
Chat
Approvals
None
What it does
Read-only
Systems
Jamf · Intune · RAG on hardware runbook
How it feels in production

An hour-by-hour walkthrough.

Priya messages Ivy: "laptop feels slow the last few days." Ivy pulls the diagnostics on Priya's device automatically: - CPU: normal (7% avg, spikes to 60%) - RAM: 92% used (34 of 36GB) — high - Disk: 91% full — high - Recent processes: Chrome (14GB), Slack (3GB), Docker Desktop (11GB) - Uptime: 34 days since reboot - Recent installs: nothing unusual Ivy responds in 20 seconds: "Your laptop is slow because of memory pressure. Chrome is holding 14GB and Docker Desktop 11GB. Two quick fixes: (1) restart your laptop to clear memory (uptime is 34 days), (2) close idle Chrome tabs — I count 87 tabs open. Also, disk is 91% full which contributes. Want me to run a disk-cleanup script that'll free ~30GB safely? Approve and I'll run it now." Priya approves. Ivy runs the cleanup, Priya restarts, laptop performance restored in 10 minutes. No ticket, no waiting, no IT-hours consumed. For issues Ivy can't diagnose from telemetry ("external monitor won't work"), Ivy walks Priya through the checklist interactively — cable check, port check, DisplayLink driver, resolution setting. Escalates to IT ops only when the checklist genuinely fails. IT sees a queue of pre-triaged tickets with the diagnostic bundle attached, not a queue of "my thing is broken."
How it works

Step by step.

  1. 01

    Pull device diagnostics on user complaint

    Immediate telemetry: CPU, RAM, disk, network, processes, uptime, recent installs, recent errors. From MDM agent + EDR + endpoint monitoring.

    Jamf · Intune · Kandji · CrowdStrike · Kolide · Fleet
  2. 02

    Diagnose from telemetry

    Common patterns matched: memory pressure, disk pressure, network signal quality, thermal throttling, process misbehaviour. Diagnosis in plain language.

    Diagnostic pattern library · Reasoning
  3. 03

    Recommend + execute safe remediation

    Restart, cleanup, driver reinstall, cache clear. Safe automations execute on user approval. Never runs disruptive actions (reformat, uninstall major apps) without escalation.

    MDM remediation scripts · User approval flow
  4. 04

    Walk through interactive checklist for unclear cases

    Peripheral / network / display issues need user interaction. Ivy walks step-by-step: check cable, check port, check driver. Interactive; branches based on user responses.

    Slack · Teams · Interactive checklist engine
  5. 05

    Escalate with pre-triaged bundle

    When checklist fails, escalate to IT ops with everything Ivy tried + diagnostic bundle + user responses. IT starts from data, not from scratch.

    Ticketing · Escalation queue · Diagnostic bundle
Systems and wiring

What you connect to make this run.

Jamf · Intune · Kandji · Kolide · Fleet

read+write

Diagnostics + remediation scripts. Read device state; write approved remediation actions.

MDM script library

read+write

Safe-remediation scripts (cache clear, disk cleanup, driver reinstall) pre-approved by IT ops. Users approve invocation.

Slack · Teams

read+write

Primary support surface. Interactive checklist walkthroughs happen in-thread.

Ticketing (Jira · Zendesk · Freshservice)

read+write

Escalation with diagnostic bundle. IT ops sees what's been tried, prevents rehashing.

What changes

Before and after, honestly.

% of hardware issues self-resolved without IT escalation
Before
20-40%
After
70-85%
Median time to resolution
Before
6-24 hours
After
10-30 minutes
IT hours per week on tier-1 hardware triage
Before
20-40 hours
After
3-8 hours
User satisfaction with hardware support
Before
3.2-3.8 / 5
After
4.5-4.8 / 5
Related

Playbooks that pair with this one.

IT

Proactive Device Health

Health telemetry that this playbook consults for reactive diagnosis.

IT

RMA & Repair

Confirmed hardware failure routes here.

IT

Network Troubleshooting

Network-specific diagnostics use the same pattern.

Frequently asked

Answers about this playbook.

What about issues that need physical access (open the laptop, swap RAM)?

Ivy escalates with the specific hardware task. IT ops schedules physical service or coordinates courier / on-site.

How does it handle issues on unmanaged personal devices?

Limited diagnostics on BYOD. Ivy asks the user to run diagnostics locally + share results. Fewer automatic fixes but same triage flow.

What about issues that recur (same user, same symptom repeatedly)?

Recurrence detected → deeper investigation triggered. Escalates to IT ops with pattern history rather than just latest instance.

Can it handle Windows + Mac + Linux equally?

Yes — per-OS diagnostic + remediation libraries. Common patterns (memory, disk, network) apply across; OS-specific patterns (Registry, Info.plist, systemd) apply where relevant.

How does it decide when to escalate vs. keep trying?

Escalates after 2-3 failed remediation attempts or if user requests. Prevents infinite loops; user always has the escape hatch to reach a human.

See it run on your data.

Free plan, no credit card. Connect the systems this playbook needs and run it against a past event first.

Book a demoMore IT playbooks