A CMDB assessment with the evidence attached.

Where your model has drifted from CSDM and why — subclass mismatches, services with no owning application, technology-derived names, orphaned offerings — each finding backed by the actual records it came from and a confidence rating, so you're triaging evidence, not opinions.

No. The assessment is read-only. She walks your live CI records and surfaces what's wrong with the modelling and how to fix it; any remediation is proposed and staged for your approval — she never edits CIs behind your back.

Completeness dashboards score fields; she reasons about the model. She looks at how services, applications and offerings actually relate in your data and flags the structural problems a field-level score can't see.

Read access to the instance. She reads the CMDB tables, relationships and classes directly, grounds every finding in a real record and cites it — nothing is inferred from a sample.

Yes. Findings come rated by confidence and impact, so you start with the modelling errors that break service mapping and reporting rather than chasing every cosmetic discrepancy.