Pressure to do something with AI
The board wants a plan, competitors are announcing new things, and what's missing is a serious read on what actually makes sense for this company.
Advice and delivery: deciding what to automate, and doing it
An honest assessment of what's actually worth automating, with priorities, ground rules for the team, and delivery to follow, if it makes sense.
Who this advisory service is for
Some companies know exactly which task they want off their plate. For them, the path is direct: our workflows and processes service, or our customer service one. This service is for everyone else — the companies that sense there's something to be gained here, without knowing where, and who are wary of anyone promising them everything.
The work is an assessment. We look at how the company runs, measure where time is going on repeated tasks, work out what's technically possible to automate today, and say in what order it's worth doing — including what's better left alone, because it's rare, because it needs judgement, or because the risk of getting it wrong outweighs the time saved.
The second half of the name is delivery. The assessment can stop here and be carried out by your own team or by another supplier, with the document serving as a brief. Or we carry on, drawing on the other services in this area and in development. The decision is yours, and it isn't tied to the assessment.
Problems this service solves
Requests for advice usually start with one of these symptoms.
Everyone using whatever they like
Half the team is already using AI tools on their own initiative, with company data, and no written rules at all.
Experiments that didn't stick
An automation or an assistant was tried before, worked fine in the demo, and was dead within two months.
Quotes that are hard to compare
Quotes come in from different suppliers, with different promises, and there's no internal way to judge them.
Nobody knows what it's worth
What's missing is the simple sum: how much time is lost today, what automating it would cost, and how long it would take to pay for itself.
An honest assessment is worth more than a pretty promise.
How we work
The assessment is deliberately short. A three-month study is out of date before anyone gets round to acting on it.
Conversations with the people doing the work
We talk to people in operations, sales and customer support, not just management. It's on the ground that you find out what actually repeats every day.
An inventory of tasks and tools
We list the repeated tasks, how often they happen and how long they take, and the tools the company is already paying for — including the ones nobody uses.
Sorting out what's worth it
Every candidate for automation is scored on time saved, technical difficulty, risk of getting it wrong and dependence on third parties. The result is an ordered list, not a catalogue of possibilities.
Usage rules for the team
We write down what can and can't be put into AI tools, which data never leaves the company, and what must always be checked by a person before it goes out to a customer.
A plan with costs and a first proof
We hand over the plan with a cost and savings estimate, and propose one small first automation — the one that proves the case before any serious investment goes in.
Guided delivery, if you want it
If we move forward together, the plan becomes a project with phases and deliverables. If we don't, the document is enough for another supplier to carry out.
What you get
The assessment is a document you use, not a presentation to file away.
A map of repeated tasks
What gets done, how often, how long it takes and who does it — often the first time the company has seen this written down.
A ranked list of opportunities
With what each one gains, what it costs and how hard it is, so you can start with the most worthwhile.
What isn't worth automating
And why. It's usually the part of the document that saves the most money.
Usage rules for the team
A short document, in plain language, on what can be used, with what data, and with what human review.
A cost and return estimate
Investment, monthly running cost and time saved, with the figures laid out plainly enough to be discussed.
A proposal for the first automation
Scope, phases and price for the first step, so you can decide with a number on the table.
Common questions about advice
The questions that reach us most often before commissioning an assessment.
What if the conclusion is that it isn't worth it?
It's a possible outcome, and we don't dodge it. That's an answer in itself: you find out your money is better spent elsewhere, with the sums already done, instead of discovering it after a failed project. The assessment is paid for the work, not for the answer coming back yes.
Are we committed to using you for delivery?
No. The document is yours, and it's written so any team can act on it, including your own. We'd rather carry on with you, of course, but an assessment that only exists to get us hired isn't really an assessment.
How long does it take?
It's deliberately short, counted in weeks rather than months, because what's being analysed changes fast. The exact timeline depends on how many people need interviewing and how many processes need mapping, and it's set out in the quote.
Do we need our data organised before we start?
No. Disorganisation is part of what we're assessing, and the first recommendation is often to tidy up before automating. If anyone tells you that you need everything in order before you can even talk to them, that's the opposite of how we work.
Does this include training the team?
The usage rules are handed over and explained. Broader hands-on training, with working sessions per department, is quoted separately, because it depends on the size and availability of the team.
Start with the assessment
Tell us how many people your company has, which tools you use and what got you thinking about this. We'll come back with the scope, timeline and price of the assessment.
Other services in Artificial Intelligence