AI advisory: honest advice in a very noisy market
Cut through the hype: where AI genuinely earns its place, where it doesn’t, and how to adopt it without the regret.
Right now you are being told you must “do something with AI.” Most of that noise is coming from people with something to sell. My AI advisory work does the opposite. I cut through the hype and tell you the truth: where AI will genuinely save you money or win you customers, where it is a distraction, and what it takes to do it safely with your data.
I am not a reseller and I do not take commission on any tool or platform. I have spent more than 20 years making technology decisions, 13 of them as a CTO protecting data sensitive enough for law enforcement, and I have built and run AI products myself. So when I tell you something is worth doing, or not, you are getting a view with no agenda behind it.
What I actually believe about AI
A few honest positions, so you know what you are getting before we talk.
- AI is a tool, not a strategy. “We need an AI strategy” is usually the wrong sentence. The right one is “here is a problem worth solving, could AI help.”
- Most businesses have a clarity problem, not an AI problem. AI applied to a messy process just gives you a faster mess.
- Start small, prove it, then scale. A small pilot tells you more than any vendor demo.
- Your data is the risk most people ignore. The part that quietly causes the damage is what happens to your information when it goes into someone else’s model.
- AI will not replace your best people. It will widen the gap between good processes and bad ones.
- If a vendor promises AI will transform your business, keep a hand on your wallet. Useful, yes. Magic, no.
Where AI earns its place, and where it is oversold
I would rather you spend money where it works and save it where it does not.
Where I see it genuinely pay off: summarising and drafting routine documents, handling repetitive customer questions, speeding up research and analysis, tidying and searching messy internal information, and taking the grind out of well-defined back-office tasks.
Where it is usually oversold: anything sold as “fully autonomous,” replacing skilled judgement, decisions where being wrong is expensive and hard to spot, and any pitch that depends on you handing over sensitive data without a clear answer on where it goes.
Why my advice is independent
I do not sell an AI platform, I do not resell licences, and I am not chasing a commission. My only incentive is to give you the right answer. When a project needs building, I can introduce people I trust, including my own agency FullyCoded, but the advice never depends on it.
I have done this, not just talked about it
I have built and operated my own AI products, not only advised on other people’s. That means you get advice from someone who knows what it actually takes to ship AI and keep it running, including the parts the demos never show you.
Frequently Asked Questions
Is AI actually worth it for a business like mine?
Sometimes yes, sometimes no, and anyone who answers that without understanding your business is guessing. That is exactly what the readiness assessment is for. I would rather tell you “not yet” than sell you something you do not need.
Everyone says we need an AI strategy. Do we?
Probably not in the way it is being sold to you. What you need is clarity on which problems are worth solving and whether AI is the right tool for them. Sometimes that adds up to a roadmap, sometimes to a single focused project.
Is our data safe if we use AI tools?
It can be, with the right guardrails, but the default settings of many tools are not safe for sensitive information. Knowing what is safe to put in, and what absolutely is not, is a core part of my advice.
Will AI replace my team?
No, and be wary of anyone selling you that. Used well, AI removes drudgery and helps good people do more of the work only they can do. The risk is not robots taking over, it is poor processes being automated badly.
What is "vibe coding" and should we be doing it?
Vibe coding means building software by prompting an AI rather than writing the code yourself. It is genuinely useful for quick prototypes and throwaway tools. It becomes risky when that code ends up running your business and nobody actually understands what it does. I will tell you honestly where the line is for you.
How do we avoid wasting money on AI?
Start small, measure honestly, and be willing to stop. Most wasted AI spend comes from committing to a big idea before testing a small one. A short pilot usually saves a long invoice.
Do you resell AI tools or take commission?
No. I have no financial relationship with any AI vendor, which is the whole point of independent advice.
Do you build the AI systems too?
My role is advisory. When something needs building, I can bring in a partner I trust, including my agency FullyCoded, while keeping the advice independent.
How long before we see results?
For a focused pilot, often weeks rather than months. Bigger changes take longer, but the approach is the same: prove value early, then build on it.
We have already started with AI and it is a mess. Can you help?
Yes, and you are not the first. Often the fix is to step back, work out what you were actually trying to achieve, stop the things that are not working, and rebuild around the one or two uses that genuinely help.