What Is an AI Consultation? What Happens and What You Get

What it actually is

A consultation maps your processes against a simple test — volume, variability, and value. How often does this run, how consistent are the steps, and what does it cost you in time or errors. Processes that score high on volume and value and low on variability are worth automating. Most are not, and saying so is part of the job.

The work is closer to an operational review than a technology demonstration. Nobody needs another walkthrough of what a language model can do. What is scarce is someone willing to sit with your specific queue, your specific handoffs, and tell you which parts are worth removing.

What happens in the session

Expect to do most of the talking. The consultant's job in the first session is to build an accurate picture of where time goes, which is information only you have.

• Where the delays are — which steps people wait on, and what they are waiting for.
• What breaks — the failures that cost you money, and how often they happen.
• Who touches what — the handoffs between people and systems, especially the undocumented ones.
• What has been tried — previous automation attempts and why they stalled.
• What the numbers are — hours per week, error rates, cost of a missed lead or a late invoice.

This is the ground we cover in an AI consultation — a diagnostic session that ends with a costed plan rather than a proposal.

What you should walk away with

A consultation that ends with nothing but enthusiasm has failed. The deliverable should be concrete enough that you could hand it to a different team and get roughly the same build.

• A prioritised list of automation candidates, scored rather than ranked by opinion.
• An estimated return for each — the hours or cost removed, against the build effort.
• A recommended sequence, because order matters more than people expect.
• The processes explicitly ruled out, with reasons.
• A baseline measurement, so the result can be proven later rather than asserted.

What it is not

It is not a sales call with a diagnostic label. If the recommendation is known before your processes have been examined, the examination was decoration.

It is also not the same as an audit. An audit describes your current state. A consultation goes further — it prices the opportunity and sequences the work. The audit is the first stage of the consultation, not a separate product.

When it is worth doing — and when it is not

The threshold is lower than most people assume. You need at least one recurring, high-volume process that consumes real time and follows consistent rules. Company size matters far less than process repetition; smaller operations often see faster returns because there is less internal process to unpick.

It is not worth doing if your processes change every month, if the work is genuinely judgement-heavy, or if the real problem is that nobody has decided who owns the process. Automation applied to an undecided process just makes the confusion faster.