hesketh.ai
One person supplies the judgment, the taste, and the relationships. A machine he built does the rest, under his hand and his name. Below are four doors and a shelf of what is coming next.
01 · the doors
Each door is at a different stage. One is open and taking clients; two are opening and taking early conversations; one is by introduction only. What is behind every door is the same: a fixed, honest piece of work, done by hand at the end, whatever the machine did in the middle.
door one · open, taking clients
This is the first of the studio's doors to open. Hesketh & Son drafts submissions for UK and Norwegian SMEs, mapped to the published scoring criteria before a word is written, and human-reviewed before it reaches you. The fee is fixed and agreed before we start, and the timeline is agreed with it. Care providers are its first focus: it reads the public record a care buyer sees, and tells you where you already score and where the gaps sit.
Hesketh & Son is UK-facing and keeps a UK correspondence address; the company behind it is registered in Norway. The full details are in the footer.
Hesketh & Son keeps its own house at heskethandson.com.
door two · opening, taking a few early conversations
For Norwegian SkatteFUNN and UK grants, the desk drafts the application and maps it to the funder's published criteria. We are opening this desk and taking a small number of early conversations while we line the work up properly. One line does not move: we draft, a named qualified person signs off the eligibility claims, and we do not file claims or give tax advice.
Tell us the grant you are eyeing and we will come back the same working day with an honest read on fit and what a real timeline looks like. No price is quoted until we know it is a clean fit for you.
door three · a small number of clients at a time
Two things live behind this door: a concierge for foreign founders setting up in Norway, and a Norway entry pack for UK firms that lost EEA procurement access after Brexit. The entry pack bundles the entity setup with a first submission-ready bid pack, drafted by Hesketh & Son, the studio's bid firm. Both are document-heavy, both drafted by the machine and finished by hand.
We are not a law firm or accountants. We draft the documents and map the process; any step that needs a regulated professional stays with someone qualified to own it, and we name which steps those are before you commit.
We take no more than two live engagements at once, because this work goes wrong when it is rushed. If we are full, we say so and give you a straight date.
Name which of the two fits and send it over. You will get a reply from a named person the same working day, with an honest read on scope and whether it is work we should take.
door four · by introduction only
This door does not take cold inquiries. It is a small advisory engagement offered to a named list. If it is for you, Dan will be in touch.
No public inquiry path for this door.
02 · about
I'm British, based in Oslo, and I've spent my career inside Nordic consumer-tech companies, most recently reMarkable, the paper-tablet company. There I designed rM Brain: a three-tier AI context architecture that the engineering department adopted as the foundational layer for all AI agentic work company-wide, serving seven departments across three platforms. Alongside it I ran the company's internal AI education, teaching marketing, sales, brand, and creative teams to get real work out of these tools.
What that work taught me is simple: AI fails quietly when it doesn't know enough, and it fails embarrassingly when it pretends to. The craft is context: giving a system the right knowledge, the right constraints, and the discipline to say "I don't know" instead of inventing.
In 2026 I left to build independently. A tender turned out to be the perfect test of everything I'd learned: a published set of scoring criteria, a pile of real evidence, and a human evaluator who has read a hundred submissions and can smell padding a page away.
03 · proof
The page itself is the first proof: the same discipline that goes into a client draft went into this page, and every claim it makes about the firm's work and record can be checked.
What it produces: source-tagged drafts, scoring-criteria maps, and the X-Ray read of a public record. No metrics, no client counts.
Philip Hesketh is the firm's Chairman. He built a £48m advertising business over 17 consecutive years of growth, wrote two Amazon No.1 bestsellers, and has keynoted for audiences at Bank of America Merrill Lynch, St. James's Place, Microsoft, and IBM. He put his name on it. Then he put his name on the firm.
04 · the machine
What you buy from any door here is a finished piece of work with a person accountable for it. The machine is the plumbing behind that: it holds context, checks its drafts against the public record, and does the heavy reading and drafting so one person can keep the judgment. Dan built it, runs it, and signs off everything that leaves it. He built it out of the craft his career was spent on: giving an AI system the right context, the right constraints, and the discipline to say when it does not know.
The proof that it works is this: the same discipline that goes into a client draft went into this page. Every claim it makes about the firm's work can be checked.
05 · the shelf
A shelf of offers, each armed with a stated trigger. A card becomes a live door when its trigger fires.
NEXT WHEN the first bid client is banked
WhatThe studio's own agent-workspace setup, packaged so another solo operator can run one like it.
WhoSolo and small-team operators who want the machine, not the service.
TriggerFirst Hesketh & Son revenue in the bank.
NEXT WHEN the first deployment-shaped client arrives
WhatA bounded, fixed-price build of one narrow internal workflow, followed by a governance and cost audit of how it runs.
WhoFirms that want one real thing automated, with someone accountable for checking it afterwards.
TriggerThe first client whose problem is a deployment, not a document.
NEXT WHEN the design conversation happens
WhatThe knowledge-architecture pattern behind the machine, built out for remote-first companies that need one.
WhoDistributed teams drowning in their own context.
TriggerA real design conversation with a company that wants it.
NEXT WHEN the record seed matures
WhatAn auditable per-employee AI-literacy record, built to satisfy the EU AI Act's literacy obligation. The record is the product; seminars are optional.
WhoEmployers who need to show, on the record, that their people are AI-literate.
TriggerThe completion-record seed matures into something demonstrable.