3 Workflows You Can Hand to AI This Week
No tools to install. No tech skills needed. Just three tasks you can stop doing manually — starting today.
Most people who say they 'use AI' open a chat window, type something vague, get a mediocre answer, and walk away unconvinced. That's not an AI problem. That's a workflow problem.
A workflow is just a repeatable task with a clear input and a useful output. When you hand one to AI properly, it stops feeling like a novelty and starts feeling like having a very fast, very patient assistant. Here are three workflows you can try this week — no accounts to configure, no integrations to build, no technical background required.
1. Email Drafting
The average professional spends around two hours a day on email. A significant chunk of that is staring at a blank reply field, trying to find the right tone for a message that is, honestly, pretty routine. That's exactly where AI earns its place.
The key is giving it enough context upfront. Don't just say "write an email declining a meeting." Say who you are, who you're writing to, what the relationship is, what you want the person to feel when they read it, and any specific detail that matters.
A strong email prompt looks like this:
'I run a small consultancy. A client has asked me to take on a project that's outside our scope. I want to decline politely, keep the relationship warm, and suggest they might come back to us for [X] in future. The client's name is Sarah. Keep it under 100 words and professional but not stiff.'
Paste what you get back into your email client, make any edits, and send. Most people find they're cutting drafting time by half or more after a week of this. The skill you're building is writing better briefs — and that's genuinely useful beyond AI too.
2. Meeting Prep
Walking into a meeting underprepared is rarely anyone's plan — it's usually just a time problem. AI can compress the prep work significantly, as long as you know what to feed it.
Before your next external meeting, gather whatever you have: the company's website copy, a LinkedIn summary, a previous email thread, notes from a last call. Paste it all into a chat window and ask AI to do the work you'd normally rush through.
A useful prompt for this: 'Here's information about the company and person I'm meeting tomorrow. Summarise what they do in two sentences, identify three things they likely care about most, and suggest two questions I could ask that would make the conversation more useful for both of us.'
You can also ask it to anticipate questions you might get asked, or to flag any gaps in what you know. This takes about five minutes and means you show up with something most people don't: a clear, calm sense of what the meeting is actually for.
If you have a transcript or recording from a tool like Otter or Teams, even better — paste the transcript and ask for a summary, decisions made, and next steps. What used to take twenty minutes of re-reading takes two.
3. Research Summarisation
This one saves professionals the most time once they see how it works. The pattern is simple: find the source material, paste it in, ask for exactly the format you need.
Say you need to understand a competitor's positioning before a pitch. You'd normally spend an hour reading their website, their case studies, maybe a few reviews. Instead, copy their key web pages, paste them into a chat window, and ask: 'Summarise this company's positioning in three bullet points, identify who their ideal customer appears to be, and note anything that looks like a potential weakness compared to a firm that emphasises [your differentiator].'
The same approach works for industry reports, long articles, policy documents, legal summaries, financial filings — anything with a lot of text and a specific thing you need to extract from it. You're not asking AI to know things it doesn't know. You're asking it to process text you've already found and give it back to you in a more useful shape.
A few things to keep in mind here. Don't paste confidential client data into a public AI tool without checking your company's policy. And treat AI summaries the same way you'd treat a colleague's summary: useful as a starting point, worth a quick sanity check before you act on anything important.
What you're building with this workflow is a habit of asking 'what do I actually need out of this information?' before you dive in. That question alone will change how you approach research, AI or not.
Pick One and Try It Today
Three workflows. Email drafting, meeting prep, research summarisation. All of them work with a free ChatGPT account or any similar tool. None of them require you to change your existing software, connect an integration, or learn anything technical.
The only thing that makes them work is doing them deliberately — giving the AI enough context, being specific about the output you want, and iterating when the first result isn't quite right. That's the whole practice.
Pick the one that costs you the most time right now. Try it on a real task today, not a practice run. Notice what you'd do differently next time. That's it. That's how the habit starts.