Aevesa

Design partner program · EU AI Act Art. 12 ready

Logging infrastructure that survives audit

Enterprises deploying agents in the EU need more than Annex IV PDFs - they need automatic logging (Art. 12) and human oversight evidence (Art. 14) that regulators can verify offline, without logging into your dashboard. Aevesa is the Prove layer: intercept high-risk actions, enforce HITL, and emit tamper-evident receipts on the Aevesa Ledger - with Assurance Autopilot running continuous cryptographic drift checks in production.

What changes when Aevesa is deployed

Moved from the homepage (P2 trim). Use this table in security reviews and procurement briefings.

Scenario Without Aevesa With Aevesa
Agent audit trail Agent actions are unaudited: no cryptographic record of what ran or why. Every action is cryptographically logged in a tamper-evident hash chain.
Gateway deny proof Gateway deny logs live in vendor-scoped Splunk. No portable offline proof. Audit evidence verified offline in under 90 seconds at verify.aevesa.com.
Compliance assembly Evidence assembled manually, often weeks of engineer time per audit. Audit evidence auto-generated: signed PDF/JSON exports on demand.
High-risk actions High-risk agent actions proceed silently. Security has no veto path. Slack Approve/Deny on every escalation with full context in one click.
Human decisions ~$4.88M average cost of an uncontrolled incident (IBM 2024). Human decisions sealed as tamper-proof activity receipts (auditor-grade records).

IBM Cost of a Data Breach 2024 · conservative $75K/intervention floor used in SAVI dashboard

Built for Art. 12 logging - not checklist compliance

Static filters and prompt guardrails do not survive contact with real agent workloads - and PDF dossiers do not prove what happened at runtime. Aevesa sits between your agents and your tools: identity-bound agents, policy on routes and tools, and the Aevesa Ledger that ties decisions to cryptography - not screenshots.

Assurance Autopilot samples your tenant ledger nightly, detects receipt digest drift, chain breaks, and mirror gaps, then signs a meta-receipt your CISO can review in the dashboard. Route → Classify → Prove: keep your gateway and AI-SPM; add Aevesa as the Prove layer your architecture diagram is missing.

  • Self-serve tiers from Sandbox through Sovereign in the app
  • Fortress: private VPC/on-prem, customer-managed keys, bespoke policy engineering
  • Direct line to product and security engineering for procurement and deployment

Fortress inquiries are reviewed individually. We respond within two business days. Champions: share the Enablement Kit with Security and Compliance before your briefing.

Request a Fortress conversation

For custom deployments only. Self-serve tiers are available at app.aevesa.com/pricing. Use your work email-we respond from our verified aevesa.com domain.