An asset register that can be trusted
ANWB runs IT assets across AWS, Oracle Cloud and Azure, alongside virtual machines on VMware and vulnerability data in Rapid7. Every platform knows its own estate. None of them knew the whole.
The CMDB was where that picture was supposed to come together: what is running, where it sits, who owns it. Keeping it current meant people updating records by hand, and hands are slower than the cloud. Auditors ask for that register. So does every conversation about cost and compliance.
What they wanted was simple to state and not simple to build. A change in any cloud shows up in the CMDB, and nobody is asked to type it in.
Make the estate report itself
A CMDB is only worth having if it is right. We did not try to make manual upkeep faster. We took the people out of the path between a change in a cloud and the record that describes it.
Agree what a configuration item is
We went through every platform holding asset data with the ANWB team and settled on one definition. Five sources, one model, before a line of code.
Build a connector per source
AWS, Oracle Cloud, Azure, VMware and Rapid7 each got their own connector, isolated so a change at one provider does not stop the other four.
Transform it on AWS
A serverless ETL layer of Lambda functions normalises what comes out of each platform into the shape the CMDB expects. No servers to keep alive.
Synchronize into TopDesk
The central repository writes into TopDesk near real time. Inserts, updates and removals all happen without a person in the loop.
Alert on what needs a decision
Not every mismatch is the machine's call. Monitoring and alerting surface the changes that need a human, and only those.
A service, not a spreadsheet
Serverless ETL on AWS
Lambda functions extract, transform and load on a schedule. Nothing to patch, nothing idling.
Five source connectors
One per platform, each replaceable on its own when a provider changes an API.
Monitoring and alerting
Failed syncs and data discrepancies arrive as alerts with an owner, so one person can run the whole thing.
An auditable CMDB
One register, current, and traceable enough to hand to an external auditor without a clean-up round first.
Serverless by default, and an eye on who runs it after us
We have been on the side of the table that inherits the platform. That is what shapes how we build one.
Cloud Native Engineering
Managed services over servers. Three engineers could build this, and one person can run it, because there is no infrastructure underneath asking for attention.
AWS partnership
The ETL runs on Lambda and the managed services around it. We are an AWS partner and build on that stack daily.
More of our work
All cases →
Would your asset register survive an audit today?
We connect the platforms that hold your configuration data, normalise it into one model, and keep the register current without manual upkeep. Then the audit is a query instead of a project.