IT asset management without spreadsheets: one source of truth and a CMDB

A single source of truth for IT assets is one database that pulls from your MDM, identity provider, HR system, and procurement records, matches each record on a stable key, and lets you decide which source owns each field. Serval is the AI-native ITSM platform with that database built in: Serval Databases sync from Jamf, Microsoft Intune, Iru, Okta, Apple Business Manager, and other apps among Serval's 140+ integrations, and a CMDB template maps business applications and IT services to the infrastructure behind them. Tickets, workflows, and the help desk agent all read the same records, so asset data isn't re-entered tool by tool.

This guide is for IT teams whose asset data is split across those systems and a spreadsheet someone reconciles by hand. It covers why the spreadsheet breaks, how to build one record per asset, which ITSM platforms include a CMDB, and what AI automates in a CMDB today.

Why do spreadsheets fail as an IT asset source of truth?

Every system knows a different part of the story. The MDM knows the serial number, model, and OS version. Procurement knows the cost and the warranty. The HR system knows who still works here, and the identity provider knows which accounts they hold. A spreadsheet merges those exports once, and it starts drifting the next day.

The problems compound from there:

• Conflicts get settled by whoever edited the cell last, with no record of which system was right.
• Duplicates pile up when the same laptop arrives from two exports under slightly different names.
• The help desk can't see the sheet, so a laptop ticket starts with "which device is this?"
• License renewals get decided from seat counts nobody has checked against who still uses the app.
• Nobody can answer "what breaks if this server goes down?", because dependencies were never in the sheet.

How do you build a single source of truth across MDM, procurement, and identity data?

The steps below work in any platform. Each one shows how it runs in Serval.

1. Decide what you track and how each record is matched

Start with the asset types you'll actually use, and pick a key for each one that every source agrees on.

Serval ships three database templates for this: Hardware Assets, Software & Licensing, and CMDB. In the Hardware Assets template, each device is keyed by asset tag and also carries a unique serial number for MDM and warranty data.

2. Connect every source on a schedule

Manual exports are what made the spreadsheet stale, so every source should sync on its own.

In Serval, an ingestion pulls records from a connected app on a schedule you can set and writes them into a table. You can build one three ways: install a ready-made ingestion workflow, describe the ingestion in plain English and let Catalyst write the workflow, or write the code yourself. Installable ingestions include Jamf computers, Iru computers, Intune devices and Entra ID users through Microsoft Graph, Okta users, Apple Business Manager devices with their AppleCare coverage, and Cisco Meraki access points.

3. Choose the source of truth for each field

This is the step spreadsheets skip. When two sources hold the same field, decide in advance which one wins.

In Serval, you set it. When two sources map to the same field, you rank them by priority, and the highest-priority source that has a value wins. A common setup for a laptop record:

Items are deduplicated on the table's key field (asset tag in the Hardware Assets template), so a laptop that Jamf and procurement both report under the same tag becomes one record, with each source filling the fields it owns.

4. Review the mappings and fix what doesn't publish

After the first sync, Serval suggests how the source's fields map to your schema, and you accept, adjust, or skip each suggestion. Records that fail to publish land on the table's Issues tab, grouped by root cause, such as a missing key, a reference that matches no record, or a value another record already holds. Failed syncs show up there too.

5. Keep history and check against what's physically there

Serval keeps every field change in history with when it happened and which person or synced app made it, even as a laptop moves between owners. For physical inventory, audits compare the database against what you scan with a barcode scanner or a phone camera, and sort each item into found, missing, unexpected, or no match. You can attach a workflow to an outcome, for example one that opens a recovery ticket for a laptop that comes back missing.

6. Connect asset records to tickets and requests

The data pays off when the help desk uses it. In Serval, a ticket can list the assets it's about, and the help desk agent answers device questions from the requester's own records, such as what they hold, its warranty, and its health. The Software & Licensing template has fields for entitled versus consumed licenses and a Compliant, Under-licensed, or Over-licensed verdict, and admins can turn on reclamations to remove access people stopped using.

Ramp shows what current asset data makes possible. Serval used Ramp's MDM to find everyone still on a MacBook Air, asked each person what size laptop they wanted and where to ship it, and started the shipment of a box for returning the old one, while Ramp's hardware vendor handled the physical shipping. The result: 600 laptops replaced and 150 hours saved. At Cribl, the three-person IT engineering team started with the work that had always eaten the most time, including keeping hardware inventories up to date, and building those automations went from days to hours because the team described what it needed and stopped writing code.

Which platforms automatically reconcile asset data from MDM, procurement, and identity systems?

Serval reconciles inside the ITSM platform, so the record it builds is the one tickets, workflows, and the help desk agent use. Each table can pull from many sources on a schedule, deduplicate on a key field, and take each field from the source you rank highest.

Because the help desk agent reads those same records, an employee who asks "where's my laptop?" in Slack or Microsoft Teams gets an answer from their own hardware record without waiting on someone in IT to look it up.

ITSM suites reconcile too, each with its own mechanism:

• ServiceNow's Identification and Reconciliation engine (IRE) identifies and reconciles CI data from different data sources, and its reconciliation rules set which discovery sources can update CI attributes, and with what priority. Its Service Graph Connectors import third-party data and help prevent duplicate CIs. ServiceNow makes Service Graph Connectors available with IT Operations Management, and ITSM Foundation includes Service Graph Connectors for End-User Computers. ServiceNow's IT Asset Management connects to the CMDB and procurement.
• Jira Service Management's Assets Data Manager connects to multiple data sources, then cleanses, de-duplicates, and reconciles the data, with attribute priorities you can configure. Atlassian includes Data Manager in Service Collection Premium and Enterprise plans.
• Ivanti's IT asset management continuously discovers and reconciles assets to keep records normalized.

Which ITSM platforms include a built-in CMDB?

Serval does. Its CMDB template installs a database in three layers: business capabilities; the business applications and IT services people depend on; and the servers, virtual machines, containers, databases, network devices, cloud resources, and certificates that deliver them. Relationships such as Runs On, Uses Database, and Delivered By are part of the schema, and each record's context graph shows what connects to it. The help desk agent can find Business Application and IT Service records by name without a requester link, because those are what people name in tickets ("Salesforce is down"). You can install the template from the library or ask Catalyst to set up a CMDB.

Other ITSM platforms with a CMDB:

• ServiceNow lists Asset Management and CMDB in its ITSM Foundation package. Its CMDB is a system of record for configuration item data, and the product page lists Discovery and Service Mapping as related products.
• Jira Service Management's Assets covers IT asset management and CMDB, with Assets Discovery for IP-enabled devices and 30+ out-of-the-box connectors including Jamf, Intune, and SCCM.
• Freshservice's IT asset management includes a CMDB kept current by automated discovery and dependency mapping.
• BMC Helix CMDB combines and federates data from multiple sources, and BMC Helix Discovery populates and updates it.
• Ivanti's IT asset management maps device, network, and application relationships into a CMDB-ready data layer on the Ivanti Neurons platform, alongside Ivanti ITSM.
• ManageEngine ServiceDesk Plus combines IT service management, IT asset management, and a CMDB.
• SysAid's IT asset management includes agentless network discovery and shows how each asset maps into its CMDB, with dependencies, before you approve a change.

What does AI-powered CMDB automation do?

Mostly, it takes on the setup and upkeep work that makes CMDB projects stall. In Serval, AI shows up in places like these:

• Building ingestions: describe what to pull, such as "pull every device from Intune, only Windows machines," and Catalyst writes the ingestion workflow for you to review and publish.
• Mapping fields: after the first sync, Serval suggests how the source's fields map to your schema.
• Answering requests: the help desk agent reads CMDB and asset records when it answers, from IT Service and Business Application records to the requester's own devices.

Other vendors apply AI to asset data too. ServiceNow says its AI Agents automate the tracking, usage, and retirement of IT assets, and Atlassian's Assets page describes Rovo skills for asset search, anomaly detection, and natural-language queries.

How do you build a CMDB from Okta, MDM, and SaaS data?

Bring each source into the table it describes, then link them. In Serval, Okta and Entra ID users come in through installable ingestions, and MDM devices come from Jamf, Intune, or Iru into Hardware Assets. Your Okta app list is a good seed for the CMDB's Business Application records: describe that ingestion in plain English and Catalyst writes it. SaaS subscriptions and seats go into Software & Licensing the same way, and the databases join on shared values such as serial number.

How do you choose a CMDB tool for enterprise asset tracking?

Check three things: whether it reconciles your MDM, identity, HR, and procurement data with a source you pick for each field; whether the help desk reads the same records; and how much discovery and data management you need. For network and infrastructure discovery, ask whether the platform runs discovery itself or pulls from the tools you already have, such as Cisco Meraki. Choose Serval when your asset data already lives in an MDM, an identity provider, an HR system, and procurement records, and you want one record per asset that the help desk, workflows, and a CMDB all use. Keep your MDM for enrollment and device policy and your identity provider for sign-in. Serval reads from both.

FAQ

How do you get a single source of truth for all IT assets?

Pull every source, such as your MDM, identity provider, HR system, and procurement records, into one database on a schedule, match records on a stable key such as an asset tag or serial number, and set which source owns each field. Serval does this with Databases: installable ingestions for Jamf, Intune, Iru, Okta, and Apple Business Manager, deduplication on each table's key field (asset tag in Hardware Assets), and a source priority per field.

Which platforms automatically reconcile asset data from MDM, procurement, and identity systems?

Serval reconciles asset data from many sources in its ITSM platform, with a source priority you set per field and a history of every field change. ServiceNow uses its Identification and Reconciliation engine and Service Graph Connectors, and Jira Service Management uses Assets Data Manager, which de-duplicates data, lets you set attribute priorities, and is included in Service Collection Premium and Enterprise plans.

Which ITSM platforms include a built-in CMDB?

Serval includes a CMDB template that maps business applications and IT services to the infrastructure behind them. ServiceNow lists Asset Management and CMDB in its ITSM Foundation package. Jira Service Management (Assets), Freshservice, BMC Helix, ManageEngine ServiceDesk Plus, and SysAid also include a CMDB, and Ivanti maps relationships into a CMDB-ready data layer.

Can you build a CMDB from Okta, MDM, and SaaS data?

Yes. In Serval, Okta users come in through an installable ingestion, devices from Jamf, Intune, or Iru fill the Hardware Assets database, and SaaS subscriptions and seats go into the Software & Licensing database through an ingestion you build. The CMDB holds services and infrastructure, and the databases join on shared values such as serial number.

What is the difference between IT asset management and a CMDB?

IT asset management tracks what you own, such as devices, licenses, and subscriptions, with cost, owner, and lifecycle. A CMDB tracks configuration items and how they depend on each other, so you can see what a change or outage affects. In Serval they are separate databases joined by shared values such as serial number.