Bespoke AI consulting · for solo operators, creators, and businesses

60–80%

less manual work, more capacity, when AI is built around the way you (or your business) actually run.

See how this works
§ 01 The problem

You’ve already noticed.

If you’re a non-technical operator (or even if you are), you’ve felt at least one of these in the last 90 days. Probably all three.

01

AI tools require skills you don’t have.

ChatGPT and Claude work — for engineers. Translating “I want to automate this” into something that actually ships requires a different kind of operator. The gap between “cool demo” and “running in production” is where most teams stall.

Result: shelfware
02

“AI consultants” ship decks, not assets.

A 50-page strategy PDF is not a working agent. After paying £20K you’re left with slides, a Notion doc, and the same problem you started with — only now you’re behind on budget and on time. Implementation is “your team’s job.”

Result: still no agents in production
03

Your ChatGPT experiments didn’t stick.

Without a sequenced playbook, single-tool experiments don’t compound. You opened a tab, prompted some things, got impressed for an afternoon — and then went back to the same workflow because nothing was integrated, owned, or measured.

Result: the work still piles up
§ 02 What you get

Outcomes you can point at.

What you walk away with depends on which engagement you bought. A diagnostic gives you a priority-scored map of every opportunity in your business and a 90-day plan you can execute. A build gives you the agents themselves: deployed, owned by your team, trained to extend. Either way there’s a tangible artifact at the end. Not another strategy deck.

60–80%
Repetitive work AI can move
What diagnosis typically surfaces: sales follow-ups, support triage, reconciliation, content drafts, hiring screens, and the long tail of admin that quietly absorbs your team’s hours.
1-of-1
Solution scope
Every solution is custom-built for the business in front of us. We don’t ship SKUs or templated frameworks, and we won’t commit to a fixed agent count on a sales call.
90d
Sequenced roadmap
Each opportunity is scored on ROI and complexity, with dependencies mapped. The plan tells you what to ship next, and why — so you stop guessing.
0×
Vendor lock-in
When the engagement ends, your team owns everything we built. You’ll have runbooks, Loom walkthroughs, and a training session so the work continues without me.
§ 03 How we work

Four ways in. One ladder.

Start where the risk fits — one focused hour, a one-week diagnostic, four dedicated implementation weeks, or an ongoing operator alongside your team. Each rung qualifies for the next. No long-term contracts to start.

§ 04 Why me

Most AI consultants haven’t shipped one.

I have. At every scale — from a billion-user platform to a one-person business. The patterns that work for both are the patterns I bring into your business.

01 · Depth

A decade inside FAANG, building AI infrastructure for billion-user products.

The work spanned agentic tooling, LLM deployment pipelines, recommendation systems, and large-scale experimentation. Most of those patterns transfer cleanly to solo-operator scale — what changes is the cost, not the architecture. The rigor I brought to global product deploys is the same rigor that shows up in your engagement.

02 · Proof

Two iOS apps. Two weeks. Solo, with agents.

FitWay (a wellness app) and Sommvi (an AI wine sommelier) are both live on the App Store. Each was built end-to-end by one person, using nothing but agents, and shipped in roughly two weeks. They have actual users, App Store reviews you can read, and paying customers. Same patterns I’d apply to a Sprint engagement, run at single-operator cost.

03 · Dogfooding

I run my own business on the same playbook.

Sales, marketing, finance, content, customer support, calendar, inbox, travel — every function inside my own operation runs on agents I built. The patterns get tested daily before any of them reach a client. I don’t theorize. I sell what I actually run.

§ 05 Cases

Three production stacks. Real numbers.

Two are my own production apps; the third is my own digital agency, run on agents. Real costs, real users, real proof of pace. Full breakdown one click away.

Role-by-role cost math → the new economics

§ 06 The process

Diagnose. Design.
Build. Run.

The same four-step rhythm — calibrated to your engagement. A diagnostic stops after Design. A build engagement carries through to Run. Where you start determines how far you go.

Step 01

Diagnose

Map every function in your business. Find the time sinks. Score the opportunities by ROI and effective cost.

Step 02

Design

Match each opportunity to the right tool stack, the right sequence, the right scope for your business. A bespoke plan — never a template.

Step 03

Build

AI systems shipped in your stack — instrumented, monitored, integrated. Sized to your business, not to a SKU. Working systems, not POCs.

Step 04

Run

Runbooks, training, on-call support. Your team operates and extends what we built — independence by handover, no vendor lock-in.

§ 07 Fit

Who this is for.

The fit isn’t really about industry or company size. It’s about how you make decisions. Almost any business has leverage AI can find; the real question is whether you’ll actually deploy what we build, or just file the report and move on.

For you, if
  • You have decision-making authority — and a budget where pace and quality are the question, not whether to spend
  • You’re done watching the AI space evolve from a distance. You want execution this quarter, not next year
  • You measure decisions by ROI you can defend, not by trends you can post about
  • You want a senior brain on your specific business — not a generic agency, not a deck
  • You can give one hour per week of senior attention during a build, and commit a designated point of contact for the duration
Not for you, if
  • You’re shopping for the cheapest “AI consultant” — I won’t be it
  • You want a 50-page strategy deck you’ll never execute
  • You expect passive income or “set and forget” automation that requires zero team time
  • You want headcount work or general operations management — that’s a real COO job, not this
  • You’re still deciding whether AI matters for your business — start with the newsletter; come back when you’ve decided
§ 08 Objections

The five questions you’re thinking.

Direct answers, no “great question” preamble.

Q.01 Will my team understand and operate the agents you ship? +

Yes — that’s the explicit design of the engagement. Every agent ships with a runbook (what it does, how to operate, when it escalates), a Loom walkthrough, and a live training session with your designated point of contact.

The Sprint includes knowledge transfer as a deliverable, not an afterthought. The Retainer includes ongoing skill-building. Independence by handover isn’t a marketing line; it’s how I’m measured.

Q.02 I have no AI tooling at all. Where do I even start? +

Start with a Strategy Call (£500) or the Audit (£3,500). The Audit ends with a recommended tool stack, sequenced subscriptions, and a quick-win section you can deploy in week one. You don’t need to know anything beforehand — that’s the entire point of the diagnostic.

Q.03 Why £3,500 for a one-week audit instead of £500 for a call? +

The call answers one question. The audit maps every function, scores every opportunity, and gives you a 90-day execution plan with a tool stack. They are different scopes for different decisions.

If you’re unsure which one to start with, take the Strategy Call. Your £500 applies as credit toward the Audit if you book within 14 days — worst case, the call pays for itself; best case, you walk away with the answer and don’t need the audit.

Q.04 What about my data? Privacy? Compliance? +

Every system we build runs in your environment, using your accounts. I don’t store, route, or train on your data. NDAs are standard before any work starts.

For regulated industries (HIPAA, SOC2, GDPR) the build is scoped around your constraints — model choice, data residency, audit logging, retention. If your stack has hard compliance requirements, we surface them in the Audit and they shape the architecture from day one.

Q.05 I’m a small business — am I too small for this? +

If you’re a non-technical operator drowning in repetitive work, you’re not too small. The Sprint pays for itself the moment it removes 8+ hours/week of manual work that compounds across the year.

The right entry point if you’re cost-conscious is a Strategy Call (£500) — one hour to confirm fit, get an honest answer on whether the math works for your business, and a recap to take back to your team.

§ 09 What to do next

Three doors. Pick yours.