Founder · Applied AI · Nottingham, UK

Building companies that run on governed AI.

Two-time founder — print, then logistics — now founder of Gardn Labs: a 20-market consumer AI app whose operations run on a governed AI agent workforce of his own design. Approval queues, earned autonomy, spend caps, audit trails.

01 · Biography

Scaled two businesses — a print group to a projected £12m turnover, and a VC-backed logistics company to £4.8m revenue with a 50+ team, a FEBE Growth 100 place and five institutional funding rounds.

Now shipping a consumer AI app in 20 markets and 13 languages, with the company’s support, moderation, content, finance and go-to-market run by 30+ governed AI agent roles. All of it built since 16 April 2026 — the day I first opened a coding tool.

Read the full biography →

02 · Selected work

Gardn Labs

A consumer AI app shipped solo in 20 markets and 13 languages — its operations run day to day by 30+ governed AI agent roles.

2026 —

Woodlark Associates

Applied-AI consultancy with Richard Ardis. Workflow-first AI adoption — blueprint, build, embed — proven on our own company first.

2026 —

We Are Fulfilment

Co-founder & CEO of a VC-backed e-commerce 3PL. £1.3m to £4.8m revenue in three years, a 50+ team, FEBE Growth 100.

2021 – 26

CCM Group

From business development to Group Managing Director. 401% growth to a projected £12m, and the pandemic PPE pivot.

– 2021

See the work in detail →

03 · The operating model

The Fragility–Autonomy Matrix

Every agent earns autonomy through clean approvals — Report, then Suggest, then Do — and every task carries a fragility class: the cost of getting it wrong. The two together make the nine-box model the whole company runs on.

Autonomy
Low fragility
Medium fragility
High fragility

Do

earned at ~30 approvals

Does the work. You read the audit trail.
Does the work — inside hard spend caps.
Does the work — with tripwires that escalate to a human.

Suggest

earned at ~20 approvals

Drafts the action. One tap to approve.
Drafts the action. You approve or edit before it moves.
Drafts the action. A named approver signs it off.

Report

every agent starts here

Watches, logs, and tells you what it sees.
Watches, and flags what looks off.
Watches only. High-stakes work always starts here.

Fragility — the cost of getting it wrong, Low → High. The highlighted row is support-01’s current lane. Swipe sideways for the full grid.

Below is one live proposal from the queue; the mechanic is the same one the company runs on.

Proposal · support-01

Reply to a customer whose plant identification came back wrong — apologise, correct it, and credit one month.

What three months of building this way looks like →

04 · Contact

Fifteen minutes, screen share — watch the approval queue run live.

Email Trent More ways in