01 / governance prospect
Konsonans AI Governance
A policy-as-code control plane for agentic systems that need traceable decisions, bounded autonomy, and operator hold points under the EU AI Act.
policy gateaudit chainhuman approval
status
building
building lane
Policy-as-code control plane for AI agents with typed action gates, identity-aware access, human approval paths, tamper-evident audit records, independently verifiable evidence packs, and Ed25519 or PKCS#11-backed manifest signing. It supports conformity work; it does not certify a deployment by itself.
finished estimate
$180k
pricing signal €25–250/mo per agent
Target studio asset value at the planned finished scope. Revenue status: pre-revenue.
Maturity is described qualitatively; no single completion percentage or derived current-value estimate is published.
What is done
The pilot path is implemented as an executable governance and evidence workflow, not only a specification.
- Typed policy gates, identity-aware access, approval queues, audit views, and operator controls are implemented across the control-plane surfaces.
- A self-contained buyer demonstration drives allow, hold, approval, and denial decisions and exports a signed real-workload evidence pack for independent verification.
- Release, demo, and deployment preflight profiles fail closed on missing requirements; evidence manifests can be signed with Ed25519 or a configured PKCS#11/HSM backend.
What is left
The remaining gates depend on the target buyer environment and independent review, not another invented compliance percentage.
- Run and preserve the validation packet inside the final buyer or operator boundary with buyer-owned identity, retention, key custody, and production integrations.
- Obtain buyer legal, security, HA/SLA, and—where applicable—notified-body review of the deployment-specific evidence and residual risks.
- Turn the verified pilot deployment into an agreed service scope, support model, and buyer-specific conformity process.
Commercial frame
The value is operational accountability: each action can be reviewed, signed, delayed, or rejected without rewriting the agent runtime.
- Target use: financial, compliance, and documentation workflows where bounded autonomy matters.
- Pricing signal: €25–250/mo per agent, matching the current product line estimate.
- Near-term sale shape: managed governance layer for teams already using AI agents in high-consequence operations.
The product is strongest when it stays close to execution: policy, approval, audit trail, and operator review in one path instead of scattered dashboards.