Edition · 15+ with squads
Govern your AI at scale. Audit every loop.
Dozens of agents and squads under a single auditable and secure framework. HALO federates the loops without losing central control.
What changes
One framework. As many loops as squads.
Federated loops + central GOVERN domain
Each squad runs its own HALO loop, but all hang from a central GOVERN domain that keeps the framework single and coherent.
Governance and risk committee
A committee reviews the Human-Only / Agent-Capable boundaries and Boundary Breaches at the organization level, not the individual level.
Templates and kill-switch per squad
Onboarding new squads with ready-made templates. Each Worker fleet has its own <60s kill-switch isolated from the rest.
AI Act compliance + security
Auditable traceability of every agent decision, aligned with the AI Act, security controls and an audit log.
Comparable metrics across squads
Signal-to-Ship measured with the same yardstick across all squads, so leadership sees who accelerates and who creates noise.
Worker fleet at scale
Dozens of Workers across code review, tests, PRs and docs in every squad, all under the same auditable and secure framework.
How it's delivered
Custom pricing. Enterprise scope.
Extended diagnosis
~€1,000
With pre-qualification. Audit of squads, loops and AI risk exposure. Deliverable: governance map and a roadmap at scale.
Custom / enterprise setup
custom
Central GOVERN domain, per-squad templates, isolated kill-switches and a comparable-metrics panel. Scope defined in the diagnosis.
Retainer · Fractional CTO
from €4,500/mo
Continuous governance, risk committee and framework evolution. Includes a monthly credit pool (1 credit = 1h = €100).
Credit model: 1 credit = 1 hour = €100. The retainer includes a monthly credit pool; scope is set after qualification.
What never changes
No matter how many squads you add, the HALO core stays the same.
The Signal → Execution → Verification loop
The human defines the signal and the boundary; agents execute; the human verifies real value.
Assistant vs Worker
Assistant: synchronous, works with you in the room. Worker: autonomous, in the background, with a <60s kill-switch.
Human-Only / Agent-Capable boundary
The explicit line of what an agent may decide alone and what always stays in human hands.
Ceremonies and metrics
The cadence that keeps the system honest, measured with Signal-to-Ship.
Bring every agent under one auditable framework.
Custom pricing. Start with a qualification call and the extended diagnosis.
Book a session