Capability statement: product for AI-augmented organisations
Decide what to build, and what to leave alone.
I help teams use AI where it clearly earns its place, and keep it out where it adds risk, cost or confusion. Over a decade in technology, a founder and consultant background, and one skill I rely on more than any tool: keeping every stakeholder informed, aligned and heard.
See how we could work togetherHow I think
I use AI every day and I like it. That is exactly why I am careful with it. The best systems I have built or guided are simple enough that a new colleague can understand them in an afternoon.
-
Simple first
The smallest thing that does the job well, and that someone else can look after in three years' time. Fewer moving parts means fewer surprises.
-
Prove it is needed
Automation has to beat the plain alternative on cost, quality or safety. Being new or impressive does not count as a reason.
-
Say it in plain words
If a non-technical colleague cannot follow the design, the design is not finished. I write and speak for the whole room, not just the engineers.
-
Keep scope honest
Automated tools can quietly stretch a project before anyone notices. I set clear boundaries early, so there is always enough clarity to adjust course.
-
Compliance is part of the design
Privacy, data handling and accountability are decided at the start, not added at the end. I know the requirements and I do the work to meet them.
The questions every idea goes through
Most ideas do not need automating, and a few need to wait. Sorting them early saves money, protects customers and keeps projects calm.
- Is the problem clear? Can the people involved describe it in one or two plain sentences?
- Are we allowed to do it this way? Is the data permitted for this use, and can we show how it was handled?
- What happens when it is wrong? Who notices, how fast, and what does it cost?
- Who is accountable? There is always a named person who signs off where it matters.
- Does it beat the simple option? If a checklist, a template or a short meeting does the job, that is the answer.
What I bring
Stakeholder management
Founders, legal, engineering, operations, customers and the people who simply have to live with the result. I keep each group informed in their own language, surface disagreements early and agree one plan everybody can stand behind. It took years to learn, and it is often what decides whether a project lands.
Product direction
Turning a vague need into a clear, ordered plan: what we are solving, for whom, how we will know it worked, and what we are deliberately not doing.
Practical AI adoption
Matching the level of automation to the job, from a simple assistant used by one person to a carefully supervised workflow across a team. I choose the lightest approach that does the job.
Compliance and risk
Privacy-aware by design: knowing what data goes where, keeping a record of decisions, adding human checkpoints and being able to explain the setup to an auditor or a customer.
Clear communication
Short, plain-language briefs, decision notes and walkthroughs that work for a board member, a support team and an engineer alike. Consultant habits, founder accountability.
Hands-on building
I can configure visual workflow tools, work with code assistants, write the code and look after the infrastructure underneath. That lets me judge what is realistic and keep things maintainable.
Maintainability
Well-documented, boring-in-a-good-way systems with few dependencies, so they keep working long after the people who built them have moved on.
How a project runs
The same five steps apply whatever tools are in fashion. Stakeholders are involved at every one of them, not just at the end.
-
1Listen
Talk to the people who do the work and the people who carry the risk. Most real requirements are hiding in these conversations.
-
2Map
Draw today's process in plain words on a single page, so everyone, technical or not, is looking at the same picture.
-
3Question
Put each idea through the questions above. Agree what to automate, what to simplify and what to park, and why.
-
4Build small
Start with the smallest version that proves its value. Easy to explain, easy to switch off and easy to hand over.
-
5Keep it running
Document it, check it regularly and review it with the people who rely on it. Calm and dependable is the goal.
Hands-on, from screen to server
Tools change every year. The shape of good work does not: connect the pieces, check the output and put a person in charge wherever it matters.
Background
More than ten years in technology, as a founder, a consultant and a hands-on builder.
Founder
I have been accountable for outcomes, budgets and customers. That is where I learned to prioritise ruthlessly, to say no kindly and to keep every stakeholder in the loop before a surprise becomes a problem.
Consultant
I have walked into unfamiliar organisations, listened first and made complicated things clear. That is where I learned to bring very different groups of people to one agreed decision.
I stay closely informed on security in the age of AI, because the risks move faster than the headlines. It keeps my advice grounded and my recommendations careful.
Where I am most useful
In teams that want AI to be useful rather than impressive.
- Organisations adopting AI that need a clear head on what to do first, what to leave and why.
- Teams working with personal, regulated or commercially sensitive information, where being able to explain and defend the setup matters.
- Groups that need someone to sit between technical and non-technical colleagues and keep every stakeholder aligned.
- Products that should still be easy to run, change and hand over in five years' time.