Deterministic by design

The same configuration, evaluated twice, gives the same result twice. That is what makes a finding defensible.

What it means for an audit

The first question an auditor asks about a finding is how it was produced. The second is whether it would be produced again. A system whose answer varies from one day to the next cannot support either.

Each rule is a boolean check on the properties of one resource, written by a person, versioned under an identifier and linked to a named requirement. Given the same configuration it returns the same result, and the record states which rule version produced it.

How the rule engine works

Written by people

Each rule is written by a person and linked to a named requirement. Mappings are reviewed and versioned. Rules are not generated or inferred.

Boolean

Each check is a boolean test of the property. If the property cannot be read, the result is recorded as not assessed. Expected and actual values are stored.

Versioned

Every record names the rule version that produced it.

Reproducible

437 rules, applied the same way on every scan. No sampling and no probability.

No statistical models in the evidence path

No pattern recognition, probabilistic scoring or trained model sits between your Azure configuration and the evidence record. The path from resource property to audit artefact is boolean logic that can be read end to end.

A probabilistic component in your compliance process would be one more system to govern, document and explain. If we ever add a capability we did not build, we will say whose it is.

Contact

Demonstrations are run in French or English, in thirty minutes, by the people who build the platform.

EqualisTech OpsReg Limited
The Black Church, St. Mary's Place
Dublin 7, D07 P4AX, Ireland
contact@equalistech.com
Request a demonstration (opens in a new tab)