Do we have to go through the levels one after the other?

The levels build on each other: an agent that is meant to act autonomously needs access to the systems and reliable data.

At which level do most companies start?

Almost every company runs AI pilots today, but only a few of them change how the company works. The Impact Check shows exactly where your company stands.

How do we know we have moved up a level?

By your KPIs and by how work is divided. When the first KPIs move, you have reached stage 2. When agents act on their own within approved limits, stage 3. When the organization steers core processes through goals, rules and KPIs, stage 4.

How do we stay in control at the higher levels?

Approvals shift with each stage. At stage 2, a person approves every proposal; at stage 3, only exceptions and cases outside the defined limits. Every step remains traceable, and the data stays within your company.

Maturity model

Four levels to the agentic enterprise

Know where you stand. Plan where you are heading.

Hardly any company starts from zero, and hardly any is already at the top. The maturity model shows where your company stands in the use of agents and which next step will have the greatest impact.

Request an Impact Check The four levels

The four maturity levels

With every level, agents take on more responsibility

From level to level, the organization shifts its attention from routine work to goals, rules and KPIs. Each level comes with the hurdle that most often holds companies there.

Stage 1
Accessible
Systems and data are accessible via APIs and MCP servers. Agents can read and transact. What holds companies here: agents cannot reach the ERP, shop and internal services, or customers, products, prices and inventory sit in sources that each hold their own version of the truth.
Stage 2
Assisted
Agents prepare and recommend, people decide. The first KPIs start to move. What holds companies here: agents speed up individual steps, while the process behind them keeps running in silos and handovers go through email and Excel.
Stage 3
Autonomous
Agents act within approved boundaries, for example in order entry, reordering or service. What holds companies here: approvals, operations and measurement are unresolved. Agents may only act on their own once it is clear who runs them, who approves and how impact is measured.
Stage 4
Agent-led
Core processes run through agents. The organization leads through goals, rules and KPIs. What holds companies here: the target model ends at department boundaries. The leap succeeds when processes are designed agentically across sales, service and logistics.

In the Impact Check, we determine which level your company is at.

More about the Impact Check

How we determine your level

Your maturity level as the basis for the next leap

In the Impact Check, we assess your company using an evaluation matrix and identify the problem areas along the four hurdles. This shows which agent will deliver impact first at your level. We meet you at your level and plan the next leap.

ProcessesWhere do workflows run in silos, and which process delivers impact fastest with agents?
DataWhere are uniform definitions of customer, product and price missing?
System accessWhich systems can agents reach today, and with which permissions?
OperationsWho runs agents, who approves, and how is impact measured?

Request an Impact Check

FAQ

What executive teams ask about maturity

: The levels build on each other: an agent that is meant to act autonomously needs access to the systems and reliable data.
: Almost every company runs AI pilots today, but only a few of them change how the company works. The Impact Check shows exactly where your company stands.
: By your KPIs and by how work is divided. When the first KPIs move, you have reached stage 2. When agents act on their own within approved limits, stage 3. When the organization steers core processes through goals, rules and KPIs, stage 4.
: Approvals shift with each stage. At stage 2, a person approves every proposal; at stage 3, only exceptions and cases outside the defined limits. Every step remains traceable, and the data stays within your company.