Independent advice for New Zealand government agencies and councils: find out what AI tools tell people about your services, where they get it wrong, and what to change so the answers are accurate.
I've led findability work at MBIE and user experience design at Health New Zealand.
Agencies are being asked to use more AI, and people still need answers they can trust. I help with both.
AI project consulting
You're choosing, scoping or developing an AI tool.
Decide where AI is useful, compare platforms on your own content, and test the answers before people rely on them. I advise, specify and test; your team or vendor builds.
A single piece of work, or ongoing alongside your team.
People or AI tools get the wrong answer about your services.
I check how your organisation shows up wherever people look for answers: Google, AI assistants such as ChatGPT and Copilot, and the other sites that quote you. You get a prioritised action plan that tells your web team, IT and vendors what to change first.
Showing up in AI answersChecking what Google, ChatGPT and Claude say from a site's pages, and improving the pages they draw on. The findability audit checks it for one site.
Getting AI into everyday useWhere it's worth using, which platforms to choose, and how to move from a pilot into everyday work.
Implementing AI featuresRequirements, evaluation and answer testing for search, answer summaries and assistants.
Smarter site searchSearch roadmaps, platform evaluation and requirements, from better matching to cited answers.
An AI model answers from Miro's background and the projects on this site. Your questions go to the model and may be reviewed to improve the assistant, so leave out anything personal (privacy). For anything else, get in touch.
Try asking
How this answer was made
Ask a question to see which model answers, how long it takes, and the tokens it uses.