Skip to content

Identity for machines

Authenticate the systems behind your systems.

Cryptographic identities for services, workloads, containers, devices and the automated processes that run your business.

The situation

What this usually looks like

Three things we are describing, not three things we are offering. If two of them match the system you have, you are reading the right page.

  1. 01

    A shared API key is in a dozen repositories and nobody can rotate it safely.

  2. 02

    You cannot answer which service called an internal API last Tuesday.

  3. 03

    Device fleets authenticate with credentials that never expire.

The approach

How this is addressed.

What the platform does about the situation above, layer by layer, before any service names are involved.

Short-lived beats long-lived
Replace shared static secrets with workload identities and short-lived credentials bound to an environment, a tenant and a deployment.
An inventory you can actually query
Keep a record of which machine principals exist, which keys they hold, when those keys expire and who is accountable for each one.
Devices from bootstrap to retirement
Enroll devices with identity certificates, support hardware-backed keys where the device provides them, and retire credentials deliberately at end of life.

The services involved

What this is built from.

A solution is a reading path through the catalogue, not a separate product. These are the services it draws on — each one with its own page, scope and integration detail.

Every service above establishes who or what is making a request. Deciding whether that authenticated principal may perform a particular action is authorization, and it stays with the consuming application or a separate authorization layer. Each service page states where it stops.

Identity for machines

Where this starts.

Create a development tenant and connect one application, or talk the architecture through first — tenancy, evidence, credential lifecycle, and which parts stay in your own stack.

Reading first is fine too: the developer portal covers APIs, SDKs and the standards each flow follows, and pricing sets out the packages and what is metered — the figures themselves are agreed with our team.