Free assessment · 2 minutes
H1VE Swarm Index
Where does your team stand on AI adoption?
A two-axis model to measure velocity and governance in AI development.
The concept
AI adoption maturity has two axes
There are two axes. V (Velocity): how the team operates with agents today. G (Governance): what sustains that operation. The core diagnosis is the gap between them.
Velocity > governance
Debt piling up: code no one can explain, decisions with no owner, risk that only surfaces in production.
Governance > velocity
Idle capacity: the process can take more AI than the team uses. There's velocity on the table, with no added risk.
V axis · Velocity
Velocity: how the team operates with agents today
Blocked
AI locked down by security and process. Usage hidden or nonexistent. No approved infrastructure.
Assisted
One dev, one agent. Synchronous work: you watch while the AI works and review almost everything by hand.
Parallel
One dev orchestrates several agents at once. The AI verifies its own work before you see it. You review diffs, not keystrokes.
Supervised autonomy
The AI writes almost all the code. Continuous loops run in the background. The human decides by exception.
AI-native
Steering by intent. Most agents are triggered by other agents. Monitoring by exception.
G axis · Governance
AI development governance: what sustains the operation
No trace
No one knows what's AI code and what's human. Decisions happen in chat and evaporate.
Informal review
Someone looks at the code before merge, but with no defined criteria. Quality depends on who reviews.
Defined gates
There are formal decision points (approved spec, mandatory review), but they cover only part of the flow.
Systematic governance
All work goes through spec, independent gates and owner review. Partial traceability.
Auditable end to end
Every decision has an owner, date and basis on record. Any delivery can be audited: what the AI generated, who approved it, on what basis.
The gap in practice
Example: V2 · G0 means stage-2 delivery sustained by stage-0 governance.
The gap between the axes defines the narrative of the result. There are five classes:
Debt zone
gap ≥ 2Your team delivers at V2 velocity, sustained by G0 governance. Every sprint in that gap piles up invisible debt: code no one can explain, decisions with no owner, risk that only surfaces in production.
Frontier
gap = 1Velocity one stage ahead of governance. This is the exact moment to close the loop, before scaling agents, not after.
Balanced start
gap = 0, both ≤ 1Balance, but at small scale. The next stage of velocity is only safe if governance climbs with it.
Governed swarm
gap = 0, both ≥ 2Velocity and governance moving together: the pattern H1VE stands for. The challenge now is holding the loop as the swarm grows.
Idle capacity
gap ≤ −1Governance ahead of velocity: the process can take more AI than the team uses. There's velocity on the table, with no added risk.
The adoption program
From diagnosis to the next stage, in 3 steps.
A binary goal, verifiable by the engagement: your team moves from V1·G0 to V2·G2. No pricing on the page: talk to us.
01
Diagnosis
Swarm Index + interviews with the team; a real map of where velocity and governance stand.
02
Loop design
Gates, roles, verification and traceability tailored to the next stage.
03
Rollout
The team operating at the new stage, with H1VE Flow as the reference tool.
Adoption-stage model inspired by the work of Boris Cherny (Anthropic), extended with the H1VE governance axis.
Next steps
AI engineering team assessment: where to go next
The Swarm Index is the starting point. Go deeper into the method, see how H1VE builds H1VE, or return to the overview.