Question 1

What is Mnemom?

Accepted Answer

Mnemom is the trust plane for the agentic internet. Every agent gets a live, cryptographic Trust Rating — composed of alignment verification, behavioral drift detection, and Ed25519-signed proofs — so your team can let agents move and still prove they stayed in bounds. Think of it as a credit score for autonomous action.

Question 2

How is this different from monitoring tools like Datadog?

Accepted Answer

Traditional monitoring tracks metrics — uptime, latency, error rates. Mnemom tracks behavioral integrity. Did the agent follow its alignment card? Did its decision-making drift from its stated values? Every check produces a cryptographic certificate, not just a log line. You can prove compliance, not just report on it.

Question 3

What does 'cryptographic proof' actually mean?

Accepted Answer

Every integrity check is signed with Ed25519 keys, chained with SHA-256 hash chains, and anchored in Merkle trees. This means results are tamper-evident and independently verifiable — anyone can confirm a check result is authentic without trusting Mnemom's servers. For high-stakes decisions, ZK-STARK proofs let you verify verdicts without revealing the underlying data.

Question 4

Do I need to change my AI agent's code?

Accepted Answer

Minimal changes. Install the SDK (pip or npm), define an alignment card describing your agent's intended behavior, and call the checkpoint function at decision points. Most integrations take under 5 minutes. The SDK handles signing, chaining, and proof generation automatically.

Question 5

Is Mnemom open source?

Accepted Answer

Yes. The Agent Integrity Protocol (AIP) and Agent Alignment Protocol (AAP) are fully open source under Apache 2.0. The core verification logic runs client-side. Mnemom's commercial offering provides managed infrastructure: hosted gateway, trace storage, dashboards, fleet management, and compliance features.

Question 6

How does pricing work?

Accepted Answer

Free tier gives you 100 checks/day with Ed25519 signing and 7-day retention — enough to prototype and test. Developer tier is pay-as-you-go at $0.01/check with ZK proofs and 30-day retention. Team tier is $79/month with 15,000 included checks, RBAC, and 90-day retention. Enterprise gets custom pricing with unlimited checks, SSO/SAML, and self-hosted options.

Question 7

What compliance standards does Mnemom support?

Accepted Answer

Mnemom is built with EU AI Act compliance in mind. Every integrity check generates an audit trail with cryptographic attestation, mapping to Article 50 transparency requirements. The system produces exportable compliance bundles including alignment cards, integrity checkpoints, and certificates. Enterprise plans include dedicated compliance tooling.

Question 8

Can I self-host?

Accepted Answer

Yes, on the Enterprise tier. You can run the full Mnemom stack on your own infrastructure, including air-gapped environments. The core protocols are open source, so you always have the option to run verification locally. Contact our sales team to discuss self-hosted deployment.

Votre agent a pris une décision. Pouvez-vous prouver pourquoi ?

Sachez ce que vos agents font réellement. Avant que le conseil ne demande.

Une Safe House autour de chaque agent.

Chaque décision, signée cryptographiquement. Pour votre conseil, vos auditeurs, vos régulateurs.

Voyez les preuves.