Insights · AI-native workflows

The real unit of AI transformation is the workflow, not the chatbot.

The short answer

A chatbot answers a question. A workflow produces a finished piece of work. Most AI programmes stall because they insert a model into one step of an unchanged process, so the process still ends where it always ended: with a person assembling the output. Transform the workflow — its sequence, its handoffs, its definition of done — and the model becomes one component of a different way of working rather than a faster version of the old one.

The question is not which model. It is which finished artefact the work produces, and who signs it.

Every organisation we work with has a list of AI pilots. Most of the lists look the same: a summariser for meeting notes, a drafting assistant for proposals, a search box over the policy library. Individually, each pilot works. Collectively, they change very little. The work still takes the same number of days, because the pilot removed a few minutes from a step in the middle and left the sequence around it untouched.

This is the most common and most expensive misunderstanding in AI programmes, and it is worth naming precisely. A step is one action inside a process. A workflow is the whole sequence from request to finished, accepted output, including the handoffs and the definition of done. Automating a step optimises an action. Redesigning a workflow changes what the organisation is able to produce.

Step versus workflow, with an example

Take a commercial team that produces a monthly pricing pack. The old workflow: three analysts pull data, build a model, write commentary, a manager reviews, a director approves, the pack is circulated. It takes nine working days and two rounds of rework.

Two ways to use the same model in the same process
Step automationWorkflow redesign
What changesCommentary is drafted by a modelThe data pull, model build and commentary become one generated artefact
SequenceUnchangedReview moves before assembly, not after
HandoffsStill threeOne reviewer, one approver
Definition of doneSameModel output must cite every figure
Effect on elapsed timeMarginalMaterial, because waiting time was removed

The right-hand column is not a better prompt. It is a different process in which the model is one participant. Crucially, it also changed the definition of done: every figure in the output must be traceable to a source row, because the reviewer can no longer assume a human assembled it.

How to tell which one you are actually doing

Three questions, asked about any AI initiative, will separate a step from a workflow. We use them before a scoping call, and they routinely change what the client asks for.

  1. What is the finished artefact, and who accepts it? If the answer is vague or plural, you are automating a step.
  2. How many hands touch it between request and acceptance? If that number is unchanged after the initiative, you have optimised a step.
  3. If the model were removed tomorrow, would the process still work? If yes, it was a step. If the process was rebuilt around the model and cannot revert without loss, you changed the workflow.
A usefully uncomfortable test

Measure elapsed calendar days from request to acceptance, not hours saved in a step. Hours saved flatter a pilot; elapsed days change the business.

What a workflow redesign actually changes

Four things change together, and skipping any one of them causes the pilot to quietly revert to the old process after three months.

  • The sequence. Work that used to happen in parallel now happens once; work that used to happen last now happens first, because a model is cheap and a reviewer is not.
  • The handoff count. Every handoff is a place where work waits and context is lost. A redesigned workflow typically removes one or two, not because the people were redundant but because the reason for the handoff disappeared.
  • The definition of done. New artefacts need new acceptance rules. If a human did not assemble the output, the reviewer needs provenance, not just plausibility.
  • The exception path. What happens when the model is wrong, or when the request is unusual? A workflow without a designed exception path fails loudly in front of a client.

The exception path is the part teams underestimate. In a human process, exceptions are handled by the person who notices them. In a rebuilt workflow, the exception has to be routed explicitly, or it becomes a silent error that surfaces months later in a number nobody can explain.

Who signs the work

There is one question that decides whether an AI workflow is trustworthy: when the output is wrong, whose name is on it? If the honest answer is “the model’s”, the workflow is not ready to carry consequential decisions, and no amount of accuracy improvement will make it so. Accountability does not degrade gracefully into a language model.

In practice this means every AI-native workflow we install has a named human owner who accepts the output, and an evidence trail that lets that owner check a sample without redoing the work. The owner is not reviewing everything. The owner is able to review anything.

The model does not remove the need for judgement. It moves judgement to a different place in the process, and makes it cheaper to exercise — if you designed the workflow to allow it.

Field note, AI-native workflow mission

This is also why the weakest AI programmes are the ones that start with a tool and look for a use. Starting from the finished artefact, the owner and the exception path forces the hard questions early, when they are still cheap to answer.

Sources and further reading

  1. Our own observation, stated as such: the step-versus-workflow distinction and the three diagnostic questions come from field work with commercial and operations teams. This is experience, not a controlled study.
  2. Evidence class: qualitative. We have not measured a population of AI programmes, and we do not present the time effects above as averages. They are the shape of change we observe, not a benchmark.
Stanley Otieno

Stanley Otieno

Instrumentation & Telemetry Lead

Questions

Questions people ask about this.

Should we start with a pilot or a full workflow?

Start with one complete workflow that has a clear finished artefact and one accountable owner. A pilot that covers a step will succeed locally and change nothing globally, which is the pattern most programmes are stuck in.

How do we know the output is good enough?

Agree the acceptance rule before the build, not after. For figures, require provenance to a source row. For prose, require a named reviewer and a sample-audit rate. If you cannot state the rule, the workflow is not yet designed.

Does this mean fewer people?

Not necessarily, and not usually in our work. It means the same people spend less time assembling and more time deciding, which tends to raise the quality of the decision rather than reduce headcount. Where headcount does change, it should be a deliberate choice rather than a side effect.

How long does a workflow redesign take?

One workflow, properly redesigned and running with your team, typically takes about twelve weeks in our work. That includes the sequence change, the new definition of done and the exception path — the parts that a tool-only pilot skips.

Field notes

One note, when there is something worth saying.

New articles on decision cadence, AI-native workflows and commercial strategy. Written by the people who ran the mission.

Please enter a valid email address.

One email when a new note is published. No sequence, no newsletter theatre, no sharing your address.

15-minute scoping call

Rebuild one workflow
before you buy more tools.

The Orchestration Accelerator rebuilds the critical workflows in twelve weeks with your team, with the exception paths designed and the owners named.