Frame the question
Half the value is in defining what you are actually deciding and why. We help you separate the live question from the noise around it.
Strategic input on the AI decisions that will shape your business.
Advisory is ongoing decision-quality guidance: a senior perspective in the room when the choices are being made. Not a deck and a handover. A relationship that improves the quality of every consequential call.
One conversation, no commitment.
What advisory is
The right voice in the room before a big call gets made.
Most AI mistakes are decided long before they are deployed. They get baked in when the question is framed wrong, when an assumption goes unchecked, or when the option that should have been on the table never even gets named. Advisory is the practice of catching those moments early.
Surface the assumptions
Every decision rests on assumptions. We name them out loud, test the load-bearing ones, and flag the ones that would change the answer.
Widen the option set
The first two options on the table are rarely the best two. We bring the comparators, the counterfactuals, and the "do nothing" baseline that most decks leave out.
How it works
A working relationship, not a deliverable.
Advisory engagements are open-ended by design. The shape adapts to where you are: a fortnightly strategy call, a board-paper review, a structured challenge before a major spend, a post-mortem after a launch. The shared thread is access to senior judgement when you actually need it.
What you get
- A single point of senior contact who knows your business
- Scheduled review sessions on a cadence that fits your work
- Ad-hoc availability when a real decision lands
- Written record of advice given, with the reasoning attached
What it is not
- A staffing arrangement or fractional executive role
- An implementation team (that is delivery)
- A scoped engagement with a defined endpoint (that is consultancy)
- A subscription for unlimited free hours
Three doors into AI
Build your own model, own a model on your data, or apply industry LLMs. The right door depends on your business, not on the fashion.
Most boards have to make this decision without a clear picture of what each door actually involves. We wrote a short explainer, with the cost and skill bar for each path, so the choice can be made with eyes open.
Building models
Training neural networks on raw data. Real research and engineering work.
Owning a model
Fine-tuning a base model on your proprietary data. A durable moat when the data justifies it.
Applying LLMs
Using industry models, with or without RAG, to automate the work you already do. Where most business value lives.
When to call us
The moments where advisory pays for itself.
- Before a major AI spend. Vendor selection, build-versus-buy, platform commitment. The decisions that are expensive to reverse.
- When the board needs an independent view.Someone who is not the vendor, not the in-house team, and not quietly billing for whatever gets approved.
- At the framing stage of a programme. Before the strategy paper goes to committee. The point where the question itself can still be changed.
- Around regulatory or governance changes. EU AI Act, sector-specific rules, internal audit findings. Where the cost of getting it wrong is enforcement, not just rework.
- When something has gone wrong. Post-incident review, root-cause framing, decision-trail reconstruction. We have done the bad weeks too.
Ready to see where you stand?
Two minutes. No email required. Find out where AI fits your business - and where the gaps are.
Or skip ahead and start a conversation