Our customers entrust us with their most sensitive signals. We design, operate, and audit Anzenna accordingly.
We connect via OAuth with the minimum scopes required. No write access unless you explicitly grant it for remediation workflows.
We ingest what we need to reason. We do not collect message bodies, attachments, or endpoint file contents.
Your graph is partitioned in dedicated storage. Models trained on your data are yours alone, never shared.
Every agent step is logged with evidence. Every action requires human approval. Every decision is reversible.
TLS 1.3 in transit, AES-256 at rest. Customer-managed keys on the Enterprise tier.
Deploy in US, EU, or APAC. Data never crosses the boundary.
Available under NDA on request: SOC 2 Type II report · Penetration test report · Security whitepaper · DPA · Sub-processor list. trust@anzenna.ai