From cohesion to fusion

In an earlier article ‘How with AI, four process domains are merging into one intelligent ecosystem’ we described how AI brings strategy, control, execution and optimisation ever closer together. That article in particular showed what changes: process management is evolving from separate domains to a learning entity in which people and technology work together.

But such an ecosystem does not emerge overnight.

In this follow-up article, we look at the development towards it. How does cohesion between process domains actually grow? Which intermediate steps can be recognised in it? And what role does AI play at each stage - realistically, without overestimating?

Not as a technology story, but as development path for mature organising.

Not a leap, but a growth path

AI rarely turns out to be the starting point of change. Rather, it is a accelerator of what is already in place. Organisations find that the value of AI is strongly related to how mature processes, information and control are set up.
The extent to which AI can actually contribute is determined by:

How explicit processes are set up,
how coherently information is managed,
How decision-making is organised,
How feedback flows between domains.

Therefore, the ecosystem does not develop linearly, but through levels and phases. Each level has a recognisable pattern of collaboration between process domains - and a matching role for AI.

Level 1 - Fragmented

Process domains exist but operate independently of each other

In this starting point, strategy, control, execution and optimisation largely function as separate worlds. Processes are implicit and highly person-dependent. Much knowledge is in heads and routines, not in explicit agreements or structures. The organisation is mainly concerned with reacting to what presents itself.

Phase 1A - Strategy pushes, execution responds

Strategy sets direction in broad terms, but receives hardly any structural feedback from implementation. Decisions are often based on experience and intuition. Executive teams solve problems as soon as they become visible, without a view on wider effects or causes.
Optimisation occurs only in incidents: when something goes wrong, it is solved - often locally and temporarily. Information is fragmented and mainly available in hindsight. AI plays a role here at most in personal productivity, for instance in writing texts or analysing individual data sets.

Features

Strategy: guiding, but hardly fuelled by implementation
Governance: informal and incidental
Implementation: ad hoc and person-dependent
Optimisation: reactive and incident-driven
Information provision: fragmented and retrospective
AI role: individual, not process

Core mechanism: do without consistency

Phase 1B - First insight emerges

In this phase, awareness arises. Data is collected and dashboards appear. For the first time, patterns become visible: recurring disruptions, structural delays or quality problems. At the same time, consistency is lacking: data are separate from processes and require a lot of interpretation.
People see that processes influence each other, but still have limited ability to manage them. Improvements remain local and temporary. AI can support here with analysis, but lacks context and scope for action.

Features

Strategy: receives first signals from operation
Steering: gets limited visibility through reports
Implementation: remains largely ad hoc
Optimisation: identifies bottlenecks without structural assurance
Information provision: patchy and not integrated
AI role: analytical, incidental

Core mechanism: awareness of missing links

Level 2 - Structured

Coherence arises through structure and agreements

In this level, processes are explicitly designed and executed repeatably. Roles, responsibilities and definitions are established. Information management introduces frameworks, allowing information to be captured and shared more consistently. Process domains are still distinct, but no longer separate.

Phase 2A - Standardisation and control

The organisation gets a grip. Reports are reliable and form the basis for periodic steering. Deviations become visible and negotiable. AI supports analysis and automation, but decision-making remains emphatically human.

The focus is on stability: predictability, mastery and control.

Features

Strategy: receive consistent steering information
Steering: fixed dashboards and consultation moments
Implementation: standardised and predictable
Optimisation: identifies anomalies
Information provision: central and reliable
AI role: analytical and supportive

Core mechanism: stabilise and manage

Phase 2B - Leveraging feedback

Data is used more actively. Patterns lead to targeted improvement actions. Teams learn from previous results and adjust processes. Strategy periodically reviews and adjusts based on trends rather than incidents.

Coherence is now structural, but still mainly cyclic: learning and adjustment happens in fixed rhythms.

Features

Strategy: steers by trends
Control: more data-driven and dynamic
Implementation: responds faster to insights
Optimisation: cyclical and structural
Information provision: shared and integrated
AI role: predictive and advisory

Core mechanism: exploiting feedback within structure

Level 3 - Integrated

Cooperation during process execution

In this level, the boundaries between steering, execution and improvement blur. Decision-making shifts from periodic to continuous. Information is up-to-date and integrated, allowing processes to be adjusted during execution.

Phase 3A - Real-time adjustment

Processes are actively monitored and adjusted as they run. AI acts as a co-pilot interpreting real-time signals and making adjustment suggestions. People monitor frameworks and make choices.

Optimisation is no longer a separate activity, but part of the process itself.

Features

Strategy: receive frequent feedback
Control: continuous and adaptive
Implementation: adjusted in real time
Optimisation: integrated into execution
Information provision: real-time and coherent
AI role: co-pilot

Core mechanism: adjust while the process is running

Phase 3B - Learning and forecasting

Processes learn structurally from data. AI recognises patterns over time and initiates improvements within agreed frameworks. The organisation shifts from reacting to anticipating.

Strategy becomes forward-looking: scenarios and forecasts form the basis for choices.

Features

Strategy: agile and predictive
Control: scenario-driven
Implementation: partly autonomously adapted
Optimisation: continuous and cyclical
Information provision: self-learning
AI role: initiating and partly autonomous

Core mechanism: structural learning and anticipation

Level 4 - Autonomous and ecosystemic

Merging into one intelligent whole

At the highest level, process domains no longer function as separate layers, but as one coherent ecosystem. Processes steer themselves within explicit frameworks. The role of humans shifts to direction, values and supervision.

Phase 4A - Autonomous litigation

Operational and tactical decisions are made autonomously. AI adjusts processes and optimises continuously. Humans monitor governance, ethics and stability.

Features

Strategy: set by people
Control: self-learning
Implementation: autonomous
Optimisation: automatic
Information provision: acting
AI role: autonomous within frameworks

Core mechanism: autonomy within explicit frameworks

Phase 4B - Strategy as co-creation

Human and AI design strategic options together. Scenarios, simulations and effects are insightful in real time. Strategy emerges continuously and is constantly recalibrated.

Features

Strategy: co-creation human & AI
Control: fully integrated
Implementation: adaptive and autonomous
Optimisation: strategic
Information provision: anticipatory
AI role: co-creating

Core mechanism: joint intelligence

What this means for organisations

This deepening makes one point unmistakably clear:

AI is not a shortcut to mature process management.

AI reinforces what is already there. Only when process thinking, information management and information literacy mature together will there be room for autonomy and fusion.

In the following article, we make this development concrete with different growth paths: In this, AI is not a shortcut to mature process management, but reinforces what is already there. We help you, as an organisation, to determine where you are now in this growth path and show which tools and steps you can use to work specifically on further maturity.