Skip to main content

Automentic

Automentic delivers Authenticated Agentic Automation: AI agents that automate complex enterprise workflows, where every action carries a verifiable identity, a cryptographic signature and an audit record.

Traditional automation asks you to trust a script. Automentic asks you to trust nothing, and gives you the evidence instead.

The problem​

When an autonomous system does something consequential — approves an invoice, updates a customer record, closes a control — three questions follow, and most automation platforms answer none of them well:

  1. Which agent did this? Shared service accounts make every action look the same.
  2. Who authorised it? A script inherits permissions; it does not inherit accountability.
  3. How would you prove it? Application logs are mutable and rarely admissible as evidence.

How Automentic answers them​

Agent identityEvery agent is issued an X.509 certificate or a SPIFFE ID and must present it before touching a system.
Action signingEach action is signed with the agent's private key, giving non-repudiation.
Continuous verificationAgents re-authenticate at critical steps. Nothing is implicitly trusted.
The audit trailAn immutable record of who did what, on whose authority, and when.

Built on open standards​

Automentic federates with the identity provider you already run and speaks OAuth 2.0, OpenID Connect, SAML 2.0, X.509, mutual TLS, SPIFFE/SPIRE, FIDO2 and W3C Verifiable Credentials. There is no proprietary identity format to lock you in, and nothing an auditor has not seen before.

See Identity standards for what each one does and where it is used.

Where to start​

Documentation status

The Concepts, Standards and Compliance sections describe the model and are complete. The Guides and Reference sections are outlines pending the product API — every page there states exactly what it is waiting on. Nothing in this documentation describes an endpoint that does not exist.