When AI comes up with HR people and owner-managers in Ireland, the same few blockers tend to come up: 1. "We don't know what's safe to put into it" (staff data, payroll, contracts) 2. "Nobody has time to learn it properly" 3. "The tools feel built for big tech companies, not a 60-person firm" 4. "Management isn't sold on it yet" Which one is the biggest for you? Reply with the number, or tell me the one I've missed. I'll take the most common answer and write a practical, plain-English post on getting past it next week.

End of the week. Two prompts, answer either or both: ✅ One win: anything in HR or hiring that went well this week. A prompt that saved you time counts. So does "I finally got that policy signed off." ❓ One question: something you're stuck on, about AI, people or pay. Someone here has probably dealt with it. I'll reply to every one. Have a good weekend.

If you're running HR (or the whole business) at a 20–150 person company, don't start with a big AI project. Start with the boring stuff you do every week: 1. First drafts of routine emails: rota changes, interview invites, "welcome aboard" notes. Give it the facts, ask for a friendly 5-line draft, then edit. 2. Summarising long documents: paste a supplier contract or policy (remove names and personal details first) and ask: "Give me the 5 points I need to act on." 3. Turning a messy list into a checklist: onboarding, offboarding, month-end. "Turn this into a step-by-step checklist with owners and deadlines." One rule: never paste employee personal data into a free AI tool. Keep it anonymised. Which weekly admin job eats most of your time? Drop it below and I'll suggest a prompt for it 👇

After any team meeting, take your rough notes, remove names and anything personal, and paste them in with this: "Turn these notes into: (1) a 3-line summary, (2) a list of actions with an owner role and a deadline, (3) any open questions. Use plain English. Notes: [paste]" Swap names for roles ("Ops lead", "Finance") before you paste, then put the names back in afterwards. It won't be perfect, but it'll be 80% of the way there. Try it this week and tell me how it went 👇

Here's a prompt I use to tidy up interview notes in about two minutes. Paste it into ChatGPT, Copilot or Claude, then paste your notes underneath. Strip out names and anything personal first. "You are an experienced HR interviewer. Below are my raw notes from an interview for [role]. The criteria for the role are: [list 4-6 criteria]. 1. For each criterion, pull out the evidence from my notes, quoting the candidate where possible. 2. Rate each criterion 1-5 and give one sentence on why. 3. Flag anything in my notes that is opinion rather than evidence, or that has nothing to do with the criteria. 4. Suggest 2 follow-up questions for a second interview." Step 3 is the useful bit. It catches the 'great energy' and 'good fit' comments that creep into notes and can't be backed up later. The AI doesn't make the decision, you do. It just gives you a cleaner, more consistent record. Try it on your next interview and tell me below: what did it flag that you'd have missed?