An AI advisor who has to make it work
Advice is cheap when the adviser never has to make it work. I build what I recommend, so my advice tends to be smaller, more specific, and honest about what isn't ready.
What people usually ask me
- Should we buy this AI tool?
- Which of our processes could AI take over?
- Why did our AI rollout stall?
- Is it safe to connect AI to our client data?
- Somebody built something clever. How do we make it work for everyone?
Good questions, and most have short answers once someone looks at how your team actually works.
How I give advice
I look at the work first. I ask to see the instructions you gave the AI. Usually there aren't any, and that's the answer to "why did it stall".
I tell you what isn't ready. Some work can't be handed to AI yet because nobody can describe it. Saying so is more useful than a plan that won't run.
I keep it small. One process, one team, a few weeks. If the advice can't be tested that quickly, it's probably too big.
I'm not selling a tool. ChatGPT, Claude, Gemini, Copilot or what you already pay for. A clear process runs on any of them.
When advice turns into building
If you want, I build what I recommend: define the process with your team, build it on your tools, test it on real work and hand it over. That's what happened at a 25-person marketing agency, now with 22 teammates set up and 16 shared team processes, and at a software startup, where the founder runs her own AI sales pipeline.
You're not obliged to. If you'd rather take the answer and build it yourselves, that's fine.
Common questions
What does an AI advisor do?
Answers practical questions about AI at your company: what to do first, which tools to use, why something stalled, and how to do it safely.
Do you sell AI tools?
No. I work with whatever fits the job, starting with what you already pay for.
Do I have to hire you to build it?
No. You can take the advice and build it yourselves. I'm happy to do the build if you want me to.
Who do you advise?
Founders and team leads at companies of 5 to 200.