If your manager keeps telling you to "use AI better" and your stakeholders keep asking for the readout a week early, and you're wondering how you're supposed to move faster and stay just as careful, I made this workbook for you. Instead of handing you a list of rules, it starts with a one-week audit of how you use AI today, then helps you build your own guardrails from what you find, so they hold up on a busy Thursday and not only on a calm Tuesday.

What's inside

• A one-week audit of your own AI use, with a spreadsheet that works out where AI saves you time and where checking eats it back

• A way to sort your tasks into execution (AI does the doing), sparring (AI pushes back on your thinking) and blended, with a practice exercise and the guardrails that change for each

• A page on what can go into AI, covering approved tools, consent, redaction and telling stakeholders what AI did

• Stage-by-stage guardrails from planning to activation, each with ideas, worked examples and a mistake I see often

• Your keep list, warning signs, review budget and a stakeholder script for when someone asks why research isn't faster now

• End-of-stage check-ins and a study close review, so you can see whether your guardrails are working

• A team section for research leads, with a 90 minute session agenda, a team charter and a starter pace agreement

The PDF is fillable, so you can type straight into it, or print it if you prefer pen and paper.