My expertise
Services
AI systems, conversion, digital strategy and Shopify development. One person for everything your brand needs to do digitally.
AI that knows your business and does the work
Most teams treat AI like a slot machine: prompt it, hope for something usable, try again. That does not scale. I build a system around your business in four layers: a knowledge base so AI knows who you are, workflows with memory so it remembers where you were, skills so it knows how you work, and agents with MCP connections so it actually gets the work done. Portable from ChatGPT to Claude to Cursor, so every session starts with your business instead of from scratch.
-
Tailored AI Knowledge Base
A structured vault with your brand context, tone of voice, products and audience. Organised in markdown and version-controlled in Git: every change is traceable, you can roll back to an earlier version at any time and the knowledge base grows with your business without turning into a mess. Usable in any AI you pick up (Claude, ChatGPT, Cursor). No vendor lock-in, no generic answers. The AI knows who you are before you start. Optionally I host the knowledge base for you, with monthly review rounds to keep it sharp.
-
Workflow Design with memory
I design how you and your team work with AI on tasks where context over time makes the difference: ad analysis, weekly dashboards, ongoing research, client intake. The result is a system that remembers where you left off, builds on earlier decisions and gets sharper the more you use it. No reset button on every new session, no briefing you retype each time.
-
Custom AI Skills
Reusable AI instructions for your recurring tasks: campaign analysis, product copy, customer replies, content briefs. Including your own decision rules, like when to scale an ad or kill it. Set it up once, reuse forever.
-
AI Agents & Automations with MCP
This is about action in your own systems, not just text from a chat. With MCP (Model Context Protocol) I connect Claude or another AI model to your backend, your data and your tools. An agent can then run recurring work, for example translations, updating product copy or generating standard reports from your data. Your knowledge base and skills define what good output looks like and where the boundaries are.
Growth & conversion
Traffic is worthless if it does not convert. I analyse where visitors drop off, why they do not buy, and what will change that. Data and outcomes, not gut feel.
UX/CRO audits
A thorough store review: where people click, where they leave, what persuades and what does not. You get a prioritised report focused on impact.
Funnel optimisation
From landing page to checkout, I walk your full funnel and pinpoint where revenue is left on the table.
A/B testing strategy
Not random testing: hypothesis-led experiments. Which headline wins? Which checkout flow converts higher?
Customer Analysis & Targeting
Who are your most valuable customers, and how do you attract more of them? I dig into your customer data (AOV, repeat rate, lifetime value, product preferences) and pinpoint which segments actually drive your revenue. Those insights translate into sharper ad targeting, offers and content.
Content & implementation
We shape content that pulls in high-intent visitors. Landing pages, product copy, and knowledge articles aimed at the searches people make when they are ready to buy.
Marketing & strategy
A great product does not sell itself. I help with strategy and execution to get your brand in front of the right people, focused on growth.
Brand positioning
Who you are, for whom, and why someone should choose you. A sharp position that steers all your marketing.
Content marketing
Strategic content for findability and authority: blogs, landing pages, knowledge articles, grounded in SEO insight.
Data analysis & learnings
Clear view of what works and what does not. I dig into your data, extract concrete learnings, and turn them into actions. GA4, Search Console, heatmaps, whatever it takes to decide on facts.
E-commerce Development
No templates, no bloat, no app overload. I build stores that are fast, convert well, and are fully yours. All custom, all under version control.
Custom Shopify development
Fully bespoke Shopify themes in Liquid. No off-the-shelf templates holding you back: code that does exactly what your brand needs.
Managed version control & hosting
Your store code in professional version control. Every change is tested before it goes live. Something breaks? Roll back in minutes.
Performance optimisation
Slow stores cost sales. I audit your code, remove unnecessary apps and scripts, and aim for sub-two-second loads.
FAQ
Straight answers on how we work, scope and ownership.
Do I need a full engagement or can I start modular?
Everything is modular. You can start with just structured context for your AI (markdown in Git), just agents with MCP, just a CRO audit, just marketing strategy or just a store build. In the intake we pick what moves the needle for your situation, not the biggest package.
What is the difference between a free strategy call and an intake?
The strategy call is no-strings: a short check on fit, where the biggest lever is and whether it makes sense to go deeper. The intake kicks off a concrete engagement: scope, timeline, deliverables and how we work together. After the strategy call you are not committed to anything; we only schedule an intake when there is real work to scope.
Who owns my AI context, skills and agent setup afterwards?
You do. That context lives in markdown in a Git repository you can own. On the site I sometimes call it a knowledge base for brevity; I mean your own source for brand, tone of voice, products and decision rules, not the same thing as a public help centre on your website. Skills, prompts and MCP wiring are delivered the same way: reproducible and handover-friendly, not a black-box SaaS. You keep using your own AI tools. No vendor lock-in. If we ever stop working together, you keep everything we built.
Why would I want an agent if I already give my AI context?
Context makes answers match your brand and rules: who you are, how you write, what is allowed. An agent uses that same context to run steps in your systems through connected tools, for example updating your catalogue or CMS or generating reports, inside boundaries we agree up front. Without an agent you keep pasting and checking yourself; with an agent the recurring part of that work is automated. In practice you usually put context and skills in place first; agents sit on top and behave more reliably that way.
What is MCP and why do you use it for agents?
MCP stands for Model Context Protocol, an open standard that lets models like Claude talk directly to external systems (stores, databases, internal tools). Without MCP you keep copy-pasting; with MCP the model performs the step itself inside your decision rules. I use it because it is open and platform-agnostic, consistent with the rest of my stack. No lock-in to a single automation vendor.
Do you only work with Shopify?
No. Shopify is where I go deepest for bespoke storefronts and Liquid, but I am not a Shopify-only shop. For strategy, CRO and AI context your stack does not matter. For development we pick what fits: Shopify when it is the right call, or other platforms and integrations you already run. Agents and integrations connect to your backend and tools as long as there is a usable API or connector.
Can I hire you for an audit only, without implementation?
Yes. You get a report with concrete, prioritised recommendations. After that you choose: implement yourself, hand it to your team, or have me ship it in phases. No obligation to book a build or retainer straight away.
What does a collaboration cost?
It depends on scope, urgency and whether you want ongoing support. I usually work on fixed project scope or agreed blocks (retainer) when you need regular sparring or execution. After a strategy call you get a straight estimate with clear boundaries. No hidden hours and no surprise invoices for work we have not agreed.