Skip to content
H1VE

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

V0

Blocked

AI locked down by security and process. Usage hidden or nonexistent. No approved infrastructure.

V1

Assisted

One dev, one agent. Synchronous work: you watch while the AI works and review almost everything by hand.

V2

Parallel

One dev orchestrates several agents at once. The AI verifies its own work before you see it. You review diffs, not keystrokes.

V3

Supervised autonomy

The AI writes almost all the code. Continuous loops run in the background. The human decides by exception.

V4

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

G0

No trace

No one knows what's AI code and what's human. Decisions happen in chat and evaporate.

G1

Informal review

Someone looks at the code before merge, but with no defined criteria. Quality depends on who reviews.

G2

Defined gates

There are formal decision points (approved spec, mandatory review), but they cover only part of the flow.

G3

Systematic governance

All work goes through spec, independent gates and owner review. Partial traceability.

G4

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 ≥ 2

Your 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 = 1

Velocity one stage ahead of governance. This is the exact moment to close the loop, before scaling agents, not after.

Balanced start

gap = 0, both ≤ 1

Balance, but at small scale. The next stage of velocity is only safe if governance climbs with it.

Governed swarm

gap = 0, both ≥ 2

Velocity and governance moving together: the pattern H1VE stands for. The challenge now is holding the loop as the swarm grows.

Idle capacity

gap ≤ −1

Governance 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.