Which tools genuinely help, how to use them well, and where they shouldn’t be used at all. Hands-on, on your own work, and measured in hours saved rather than slides produced.
Nobody has had time to work this out properly. The tools arrived faster than anyone’s ability to assess them.
There’s an enormous amount of noise, most of it written by people selling something. Finding the two or three things that would genuinely help you is a short exercise.
Staff adopt these tools long before a policy exists. That isn’t misconduct, it’s initiative without guidance, and one clear page and a conversation usually settles it.
Almost always because it was demonstrated on somebody else’s work. People take to these tools after watching them handle the thing that is actually sitting on their own desk.
It usually can, and not with a large system. A small tool built around that single job, which you own and can maintain yourself, is a few days of work rather than a project.
A look at how the work really gets done, and which repeated tasks are worth handing over. Usually fewer than people expect, and rarely the ones they guessed.
Which subscriptions earn their keep, which overlap, and which to skip entirely. Including the annual cost once everyone has a seat.
Hands-on sessions using your actual work rather than a generic demonstration. People learn this by doing their own job with it, not by watching someone else’s.
The habits that separate genuinely useful from mildly frustrating: giving context, working in steps, knowing when to start over rather than argue with it.
The discipline almost nobody teaches. What these tools get wrong, how confidently they get it wrong, and what must always be verified before it leaves the building.
Custom assistants, document workflows and automations for the tasks you do a hundred times a month. Small, specific, and built to be maintained by you.
What may be put into which tool, what mustn’t, and how that squares with POPIA, client confidentiality and the contracts you’ve already signed. One page people will actually read.
Whether any of this is saving real time, or simply feeling modern. Agreed before we start, so there’s an honest answer at the end.
I’m not going to pretend to be a machine-learning engineer.
What I do is practical: choosing the right tools, getting people genuinely competent with them, building small things that save real hours, and keeping you out of trouble with data you shouldn’t be pasting anywhere.
One-to-one, on your own work.
Not everyone needs a team programme. Plenty of people just want a couple of focused hours with someone who uses these tools daily, applied to the work actually in front of them.
Consultants, accountants, lawyers, writers, agency owners, people running a business on their own. We sit down with your real tasks and leave with a handful of workflows you’ll use every week.
Fixed fee for workshops and scoped work, monthly for ongoing. The field moves quickly enough that the last one is genuinely useful rather than a way of keeping you on a retainer.
Your people, your actual work, hands on keyboards. Everyone leaves with three or four things they’ll use the same week, and a shared sense of what not to do.
Where it helps, which tools to buy, how they’re configured, and a usage policy written in plain language. Delivered as a short document you own.
A specific assistant, automation or document workflow for a task you repeat constantly. Built small, documented, and handed over so it isn’t dependent on me.
Just you and your own work. We find the handful of workflows that fit how you actually operate, and set them up properly before you leave.