Skip to content

Device Inventory Reconciliation

Example prompt: "Once a month, compare the laptops in our MDM with the staff directory and the asset register, flag anything that does not line up — devices assigned to leavers, devices with no owner, people on the staff list without a device on record — and send me a tidy exceptions list."

The Problem

The MDM, the asset register in a spreadsheet, and the staff directory in the HRIS all describe the same fleet of laptops, and they all drift. The MDM thinks Alex still has the MacBook because nobody factory-reset it before the courier came; the asset register lists a Dell that finance bought last quarter but nobody has enrolled; the staff directory has three contractors who never had a company laptop in the first place. The IT lead does the reconciliation once a year because pulling all three exports and aligning the columns is a half-day job, and the auditor finds five issues every time.

How GloriaMundo Solves It

We build a workflow that runs on the first working day of the month. An integration step pulls the device list from the MDM (Intune, Jamf, Kandji, or the Google admin console for ChromeOS), the asset register from Google Sheets, and the active staff list from the HRIS. A code step normalises identifiers across the three sources — serial number, asset tag, assignee email — and joins them on the assignee. A conditional step bands each row by exception type: device in MDM with assignee no longer in HRIS (leaver still in MDM); device in asset register not in MDM (missing enrolment); staff member in HRIS with role that should have a device but no device in either source (provisioning gap); device in MDM with no assignee recorded (unassigned device). An LLM step drafts a short note for each exception explaining what the issue is and the likely next action. An integration step writes the exceptions list to a Google Sheet for the IT lead and sends a single Slack summary in #it-ops with the counts by band and a link to the sheet. Glass Box preview shows the normalised join and the drafted notes before anything is sent.

Example Workflow Steps

  1. Trigger (scheduled): First working day of every month at 09:00.
  2. Step 1 (integration): Pull the device list from the MDM (Intune, Jamf, Kandji, or the Google admin console for ChromeOS) — serial number, model, last check-in date, assignee email.
  3. Step 2 (integration): Read the asset register from Google Sheets — serial number, asset tag, purchase date, assignee, location.
  4. Step 3 (integration): Read the active staff list from the HRIS — name, email, team, role, start date, last working day.
  5. Step 4 (code): Normalise identifiers across the three sources and join them on serial number, asset tag, and assignee email; classify each row by exception type.
  6. Step 5 (LLM): For each exception, draft a short note explaining what the discrepancy is and the likely next action — wipe, enrol, assign, recover.
  7. Step 6 (integration): Write the exceptions list to a Google Sheet for the IT lead with one tab per band, and post a single Slack summary in #it-ops with the counts and a link to the sheet.

Integrations Used

  • MDM (Intune, Jamf, Kandji, Google admin) — the source of truth for which devices are enrolled and who they are assigned to
  • Google Sheets — the asset register and the exceptions report
  • HRIS (BambooHR, Workday) — the active staff list used to identify leavers and unrecorded provisioning gaps
  • Slack — the morning summary in #it-ops with the counts and the link

Who This Is For

Internal IT teams managing fleets of fifty to a few hundred laptops where the MDM, the asset register, and the HRIS are all in use but no single tool joins them, and where the annual audit currently surfaces the same kinds of exceptions every year.

Time & Cost Saved

A manual three-source reconciliation is a half-day a quarter at best and never happens monthly because the friction is too high. The monthly run keeps the exceptions list short — typically a handful of items per run rather than dozens at the year-end — and the saving is partly the IT lead's time and partly the cost of unaccounted laptops, which on a fleet of a hundred at six hundred pounds depreciated value each is enough to justify the workflow on lost-asset terms alone.