Why automating the current process is the floor, not the ceiling
There are two kinds of people who can deploy an AI workflow for a client. From the outside, they look nearly identical. Both work with AI platforms. Both write prompts. Both build integrations. Both produce something that, when it launches, appears to work.
The difference shows up six months later.
The commodity offer
The first kind of practitioner — the commodity provider — builds by mirroring. Their first question is: tell me how you currently handle this, and I’ll make AI do it. They map the existing process, identify the steps, and configure the AI to follow those steps faithfully. If the customer calls about a billing issue, the AI collects the account number and describes the refund policy, just like a well-trained human agent would. If the customer calls about a lost item, the AI asks for a description and creates a report, just like the current staff process does.
This is not dishonest work. It is real implementation. And in many cases, the result is genuinely useful — faster, more consistent, available around the clock, more scalable than the human process it replaces.
But here is the ceiling of this approach: it is bounded by the current process. The AI cannot produce better outcomes than the process it is following. If the process has gaps — undefined ownership, informal workarounds, authority steps that nobody has ever formalized, failure modes that experienced staff navigate intuitively — the AI inherits those gaps. It executes the broken parts faithfully along with the functional parts. And it does so at scale and at speed, which means the broken parts produce broken outcomes more efficiently than before.
The commodity provider has delivered what they promised: AI that follows the current process. They have not changed the process. They have not asked whether the process should change. The client is getting faster and more consistent delivery of whatever they already had.
The gap that most clients can’t see
Here is the part that makes this situation complicated. Most clients cannot see what they’re missing. They know what their current process produces. They don’t know what a well-designed process could produce. They have no reference point for the gap between their as-is reality and a genuine best-class workflow.
This is not ignorance — it’s just the nature of operating inside a system. You see what the system does. You don’t easily see what it fails to do, especially when the failures are diffuse and gradual: the customer who didn’t call back after a frustrating AI experience, the refund that took three weeks because the authority path was undefined, the lost item that was never retrieved because the notification never reached the right staff member.
A commodity provider works within this blind spot. They ask the client what the process looks like and build accordingly. They never hold up a picture of what the process could look like, so the client never has the opportunity to see the distance between here and there.
The second kind of practitioner — what I would call a serious AI workflow designer — makes that gap visible before building anything.
The transformation offer
The designer’s approach looks different from the beginning. Before asking about the current process, they ask about the outcome. What is this customer actually trying to achieve? Not what step are they on, not what category does their request belong to — what condition would make their problem no longer a problem?
From that answer — the solution statement — the designer generates a picture of the ideal workflow. Not the workflow constrained by what the client currently has. The workflow that would reliably produce the desired outcome if the right systems, records, staff processes, and authority paths existed. This is the to-be model: an independent benchmark for what good looks like.
Then — and only then — the designer works with the client to understand the current reality. Where does the current process match the ideal? Where does it fall short? Where are the gaps? What would need to change in the business process for the AI to reliably deliver the to-be outcome?
This is a different conversation than the commodity conversation. The commodity conversation is: here is what you do, we will make AI do it. The transformation conversation is: here is what the ideal looks like, here is your current reality, here is the distance between them, and now we need to decide how far toward the ideal your organization is willing and able to move.
That distance — the gap between as-is and to-be — is where most of the actual value lives.
Why the value is in the gap
Consider the lost-item workflow again. A commodity implementation gives the client an AI that creates lost-item reports quickly, consistently, and at any hour. A transformation implementation gives the client an AI that creates reports — and also surfaces the question of whether those reports ever reliably lead to items being found. Who checks the reports? Who searches when a new report comes in? Who is authorized to verify a match and release an item to a customer? What is the protocol for notifying the customer when something is found?
These questions are not AI questions. They are operations questions. They are questions about how the business actually works and whether the business process is designed to deliver the outcome the customer needs. The AI is only as good as the process it sits on top of. If the process has no defined authority path for item verification, the AI cannot produce item verifications. If nobody is checking the report queue, the AI’s ability to create reports quickly is irrelevant to the customer who called about their phone.
A commodity provider builds the AI part. A workflow designer builds the AI part and addresses the process part. The client has to be willing to engage with that conversation — to look honestly at their current operations and make decisions about what needs to change. Many clients, when they understand the difference, choose the transformation path. Because the transformation path produces better outcomes, and better outcomes are what they actually wanted.
Some clients don’t want that conversation. They want the AI implementation without the process scrutiny. This is a legitimate choice, and a commodity provider serves that need. But a serious workflow designer owes it to the client to name the gap before accepting the constraint. The conversation goes: here is the ideal, here is the distance between the ideal and your current process, here is what you get if we pursue the transformation, and here is what you get if we build against the current process. You decide.
That transparency — showing the client the gap before building, rather than building inside the gap without mentioning it — is what distinguishes the two orientations. It is also what protects both parties when the AI delivers exactly what was promised and the client is still disappointed, because what was promised was faster execution of a process that wasn’t producing what they needed in the first place.
The ceiling of the commodity offer is the current process. The ceiling of the transformation offer is the best the business can operate. Those are different ceilings. Which one the client ultimately builds toward is their choice. But they can only make that choice if someone shows them both.
Since February 5, 2026, I’ve had the opportunity to study feedback from more than 18,000 complicated AI integrations running in production across 43 projects.
That work has taught me something important: AI workflow design is not just automation. It is a new discipline.
Humans do not arrive as clean tickets, categories, forms, or process steps. They arrive unfinished. They speak, the problem emerges, and the workflow designer has to understand what solution the human is actually reaching for.
I wrote this textbook to help other workflow designers think more clearly about that process.
The framework is built around five movements:
Emergence gives you the solution. Hallucination gives you the ideal. As-is discovery gives you reality. To-be selling gives you transformation. Stewardship gives you renewal.
The textbook is free to download below.