Before You Add AI, Redesign the Work.
Why enterprise AI transformation should start with processes, people and operating models — not with the technology.
What you need to know in 60 seconds.
Adding AI to a broken process can simply make a bad process operate faster.
Understand how work actually happens — not only how the process is documented.
Decide what technology should do and where human judgement creates more value.
Measure AI transformation through business outcomes, not the number of AI tools deployed.
Artificial intelligence is becoming remarkably easy to access.
Powerful models are available through APIs. AI assistants can be deployed quickly. Automation platforms can connect applications that once required significant custom development.
For a large organisation, however, access to AI is rarely the real challenge.
The difficult part is deciding where AI belongs, how work needs to change around it, which systems it needs to interact with and how people will work with it once it is deployed.
THE PRINCIPLETechnology should follow the problem. Not the other way around.
AI can improve a process. It can also accelerate a bad one.
Most organisations naturally look at AI through the lens of efficiency.
Where can we save time?
Which repetitive tasks can we automate?
Where can AI reduce manual work?
These are useful questions. But they can lead to the wrong starting point.
A customer request arrives by email.
AI could read the email, extract information, classify the request and prepare a recommendation.
But before automating anything, there is a more important question:
Why does the process work this way in the first place?
Perhaps the spreadsheet should not exist.
Perhaps the CRM and ERP should already exchange information.
Perhaps approval is required only for exceptions.
Perhaps employees are compensating for poorly integrated systems.
If we simply add AI to the existing workflow, we may successfully automate a process that should have been redesigned.
Understand the real workflow. Not just the documented one.
The process shown in a diagram is rarely the complete process taking place inside an organisation.
The official workflow.
Systems, policies, documented steps and defined responsibilities.
The workflow people actually follow.
Spreadsheets, emails, workarounds, undocumented exceptions and manual checks.
Employees create spreadsheets because the official system does not provide the information they need.
Managers introduce additional checks because of problems that occurred years ago. Information moves through email because two platforms do not communicate correctly.
Experienced employees often know how to manage exceptions that were never formally documented.
Where work begins
Where information comes from
Who makes decisions
Where delays occur
Which steps add value
Which steps are historical workarounds
The goal is not to automate everything.
Removing human involvement is not automatically a sign of progress.
Some activities are repetitive and provide very little value when performed manually.
Others require judgement, experience, context, accountability or human interaction.
Let AI handle the preparation.
- Collect information
- Analyse documents
- Identify inconsistencies
- Prepare recommendations
- Highlight unusual activity
Keep humans where judgement matters.
- Understand context
- Review recommendations
- Handle exceptions
- Manage relationships
- Make accountable decisions
The objective is therefore not to design a process with the fewest possible people.
It is to determine what technology should do and what people should do.
AI gives us an opportunity to redesign work.
Traditional automation works particularly well when a process is predictable and rules can be defined in advance.
AI expands what can be redesigned because modern systems can work with language, documents and less structured information.
“Where can we insert AI into our existing process?”
“If we designed this process today, would we design it the same way?”
The goal is not to automate the workflow you have. Design the workflow you should have.
The employee’s role should change too.
If AI can retrieve the information, organise it, summarise relevant history and highlight potential issues, the employee’s role changes.
They are no longer primarily an information collector.
Their value moves towards reviewing analysis, understanding context, handling exceptions and making the final decision.
The technology has not simply made the employee faster. It has redesigned the work around the employee.
Involve people before the technology is finished.
Employees are sometimes treated as the final stage of an AI project.
That sequence misses one of the organisation’s most valuable sources of knowledge.
The people performing the work often understand the process better than anyone else.
They know which customer requests create problems.
They know which information is normally missing.
They know which reports nobody actually uses.
They understand which exceptions require experience.
They know which parts of the official process are quietly ignored.
Their knowledge should influence the design of the AI solution, not simply its adoption afterwards.
AI has to work with the business you already have.
Large organisations rarely operate inside one application.
AI needs to operate inside that reality.
This is why integration often becomes as important as the AI itself.
An intelligent system creates much less value if employees still need to manually copy information between systems before the AI can use it.
The best AI transformation contains relatively little visible AI.
The value can come from better integration, cleaner information flows, process redesign and AI applied only where it matters.Governance belongs inside the workflow.
Experimenting with an AI assistant personally is very different from putting AI into a process involving customers, money, sensitive information or thousands of transactions.
Who is accountable for the output?
When must a person approve the result?
What information may the system access?
What happens when a case is unusual?
Can a decision be investigated later?
How is performance measured after deployment?
Governance should therefore not live outside the process as a policy document.
It needs to become part of how the workflow operates.
AI transformation is a multidisciplinary problem.
Priorities, economics and strategic objectives.
How the work actually happens every day.
Architecture, data, integration and security.
What the technology can and cannot reliably do.
The model is one component. The organisation around the model is the system.
Do not measure AI. Measure the business.
Measuring whether an AI system performs correctly is necessary.
But a technically impressive AI implementation can still create very little business value.
Did the process become faster?
Was repetitive work reduced?
Were mistakes reduced?
Did customers receive answers faster?
Did the economics improve?
Are people actually using the new workflow?
Where should a CEO start?
Start with the business
Define what the organisation is actually trying to improve.
Map the current process
Understand how work really happens, including exceptions and workarounds.
Find the friction
Identify where time, cost, complexity, manual work and risk exist.
Define the better outcome
Decide what should become faster, simpler, safer or more scalable.
Then choose the technology
Select the right combination of AI, automation, software, integration and human judgement.
Redesign the work.
AI gives businesses a new reason to question how their organisations actually operate.
Why does this process require five steps?
Why does the same information exist in three systems?
Why is this report created every week?
Why does this decision require thirty minutes of preparation?
Why are people manually moving information between systems?
Why is an experienced employee spending hours on work that requires little of their expertise?
These are transformation questions.
Most businesses will eventually have access to similar AI technologies.
The difference will be in how intelligently they redesign the organisation around them.
The objective is not simply to put AI into the workflow. The objective is to build a better way of working.
Start with the business. Then apply the technology.
Through the SHIFT™ methodology, we help organisations understand where value can be created, redesign processes, align people and technology, implement the right solutions and turn AI potential into measurable operational change.
Research behind the perspective.
The Operating Model Advantage: Why AI Winners Are Rewiring Their Organizations
Examines why AI value depends on redesigning how work and decisions happen across the organisation.
Rewired Takes: Practical People Lessons for Scaling AI Adoption
Focuses on workflow redesign, leadership, capabilities, operating models and culture.
AI at Work: Strategy Matters More Than Tools
Explores the gap between AI adoption, productivity gains and measurable organisational value.
How AI Is Helping Companies Redesign Processes
Examines AI as an opportunity for business-process redesign rather than simply another automation layer.
Artificial Intelligence Risk Management Framework
A practical framework for managing AI risk and supporting trustworthy and responsible AI.
OECD AI Principles
Principles covering trustworthy AI, human oversight, robustness, security and accountability.