Training

• Your real work Each session runs inside a task you actually do: the email, the report, the spreadsheet, the presentation. No slideware.
• Your tools Set up properly in the first session: account and data settings, what never gets pasted in, and standing instructions so the tool already knows how you work.
• Checking the output The skill that matters most. How to read what AI gives you, what it gets wrong, and how to catch it before anyone else sees it.
• Your playbook After the third session: your prompts, your rules, and the checks, written in plain language so it keeps working without me.

Someone who keeps hearing they should be using AI and wants to actually get good at it, without booking a half-day event for the whole office. If that describes three or more people on one team, the team sessions are better value.

• Pick the tool The one that fits your work and your budget, not the one with the loudest marketing.
• Set it up right Account settings, data and privacy options, and standing instructions, done together so you know what each choice means.
• Three real tasks We walk through three things you do every week, so you leave with working examples instead of theory.
• One page A record of what was set up, what each setting does, and why it’s set that way.

Most of the early mistakes with these tools happen before the first prompt: the wrong tier, the wrong data settings, no ground rules. Two hours fixes that. If you want to keep going afterward, this fee is credited toward the coaching package within 30 days.

• What the tools do And what they don’t, in plain language, without the vendor gloss.
• The questions you'll get Privacy, client data, and “did AI write this?”, with answers you can stand behind.
• What a sensible policy looks like The handful of rules that matter for a small organization, and the ones that just create paperwork.

Legal advice, and not a course on prompting either. It’s the working knowledge to make decisions about AI in your organization and to ask the right questions of the people you answer to and the people who answer to you.

• The audit Every third-party request your site makes, page by page: embeds, fonts, trackers, analytics, and what each one tells someone else about your visitors.
• The fixes Embeds converted to click-to-load, so YouTube and social widgets load only when a visitor chooses them. Analytics moved to a cookieless option if you’re on one that needs a consent banner.
• The report What was loading, what loads now, and what it means, written for whoever answers the privacy questions.

I’ve done this pass on my own sites: every third-party embed converted to click-to-load, with the requests on each page checked before and after.

• The audit Heading structure, contrast in light and dark modes, keyboard navigation, alt text, and how the site reads to a screen reader.
• The fixes Applied directly where I have access to the site, or written as specific instructions for whoever maintains it.
• The report What was wrong, what’s fixed, and what to keep doing as the site changes.

This site passes an automated accessibility scan on every release, and I’ve run the same phased pass (headings, contrast in both colour modes, semantics) across my own multi-page sites.

• The read You send the AI-drafted newsletter, post, or letter before it goes out. I read it the way your sharpest reader would.
• The notes What’s wrong, what’s off-key, what the tool made up, and the fix for each, back within two business days.
• The pattern Over a few months, the notes repeat less, because the notes are also teaching whoever writes the drafts.

AI drafts are fast and mostly fine, and “mostly” is the problem: the errors that remain are confident ones. A second read from someone who edits this material every day costs less than fixing a mistake after it’s gone out under your organization’s name.