Sarghaby Araghatta
Home / How coverage is measured

Coverage you can measure, and never 100%.

Sargha counts the machines and accounts it knows about, says how many it can actually see AI on, and lists the rest as named gaps.

SensorSees the host directlyConnectorSees its AI useTelemetry onlyIn no inventoryKnown but blindListed, cannot see AIUnknownCannot be matchedSensorSees the host directlyConnectorSees its AI useTelemetry onlyIn no inventoryKnown but blindListed, cannot see AIUnknownCannot be matched
Every known entity lands in exactly one tier. The best source wins.

One honest number

Covered means a sensor or a connector sees its AI. In a worked example with round numbers (illustrative, not customer data), 620 sensor and 210 connector entities out of 1,000 known is 83.0%. The 110 known but blind are a named list.

A sensor that goes quiet stops counting

A sensor that has not checked in recently no longer counts as covered, and the machine moves to the blind tier.

Your estate is counted, not guessed

Identity providers, device management, network inventories, cloud accounts, Kubernetes nodes and a CMDB or CSV import say what exists. Add a source and the number may fall, which is honest news.

Each class on its own

Endpoints, servers, phones, network devices, IoT and OT, storage, SaaS, cloud accounts, identities and AI gateways are scored separately, so a strong endpoint score cannot hide a weak one.

A fix for every gap

Each blind entity is listed with the source that would close it, so the next step is a decision and not a debate.

Why never 100%

Anything in no source is invisible. Network evidence covers only its own network. A DNS lookup is not proof of use. Some devices cannot run a sensor. Inventory is not usage.

How it works

Step by step.

  1. Count what exists

    Identity providers, device management, network inventories, cloud accounts and imports say what is out there.

  2. Place each entity in one tier

    The best source wins. A machine with a sensor and a device-management record counts once, as a sensor.

  3. Work down the gap list

    Each blind entity is listed with the source that would close the gap.

Questions

Common questions.

Does adding a source lower my score?
It can. The total may grow, which means you now know about more that you cannot yet see. That is honest news, not a regression.
Is demo data counted?
No. Demo data is excluded from every number.
What you need
  • Connect an identity provider and a device-management or CMDB import first. The score and the blind-spot list appear as soon as they sync.
Get started

See the AI on your fleet in about a week.

Run the sensor across your estate and receive an AI risk report and an AI Bill of Materials. The assessment is free and carries no commitment.

Request a free assessment → Book a 30-minute call