One trusted CMDB.
Finally reconciled.
A hybrid CMDB built around AI reconciliation. Lightweight collectors discover vCenter, Kubernetes and F5 — then an Identity Engine turns duplicate chaos into trusted golden records.
Hybrid collectors,
where your data lives
Lightweight collectors run right next to your infrastructure — on-prem for vCenter, Kubernetes and F5, with a generic Cloud Connector for major providers on the way.
- vCenter via vSphere SDK — VMs, hosts, clusters, datastores
- Kubernetes — nodes, namespaces, pods, deployments, services
- F5 BIG-IP over SNMP — VIPs, pools, members, iRules
- Cloud Connector for GCP / AWS / Azure — coming soon
One CI. Not
five duplicates.
The same server shows up in vCenter, Kubernetes and your spreadsheet — under three different names. The Identity Engine correlates them into a single golden record with a confidence score you can audit.
- Weighted multi-signal scorer: hostname, IP, BIOS UUID, environment, graph
- Deterministic match first, similarity + graph correlation for the rest
- Simulation mode — preview every merge before you commit
- Low-confidence pairs route to a human review queue
A Quality Score
on every record
Every CI earns a 0–100 score from completeness, ownership, recency and relationship coverage. Finally, a way to see — at a glance — exactly what you can trust.
- Completeness — are the key fields populated?
- Ownership — does the CI have an accountable owner?
- Recency — how fresh is the last discovery?
- Relationships — is it wired into the service graph?
Two cost lenses on
the same estate
Discovery already knows your physical cores and every VM's allocation. FinOps turns that into two honest views — the licensing footprint to budget for, and what each team should be charged back.
- Costs computed from discovered hardware — you supply the rates
- Rightsizing & reclaim candidates flagged for review, never auto-actioned
- Chargeback rolls up by owner, business unit and cluster
Ask in plain
English (or Portuguese)
Claude translates a natural-language question into a transparent query plan — you see the filters before anything runs, then export the result straight to CSV.
- Query plan shown before execution — no black box
- Post-filters for ratios like free-space % or vCPU thresholds
- One-click CSV and ServiceNow Import Set preview
"ciClass": "NodeCI",
"filters": { "osFamily": "linux", "powerState": "poweredOn" },
"postFilter": "memoryMb > 8192"
}
Ready when you are
Hybrid collectors, AI reconciliation, Quality Scores, and VMware FinOps — that's what CMDBHub does. Get the free Community Edition or book a 15-minute live demo with your own data.