HumanGrow helps leaders and organisations grow people, judgement and governance at the pace AI is growing.
The arc
Stage one
Eighteen years shipping enterprise platform capability, the last two of them on AI itself: a Copilot programme across more than forty countries, a governed runtime for internally built applications, agent architectures with their boundaries enforced in the database rather than the prompt. The work at this stage is to make the thing exist and reach the people who need it.
Stage two
Delivering at that scale exposed the real blocker, and it was not model quality. Nobody could say what a given AI capability was allowed to touch, or on what evidence it had been trusted. So ENGRAM was built to answer those two questions before a capability runs. The work at this stage is to make delivery defensible after the fact, to an auditor or a board.
Stage three
The loop closes when governance stops being a layer wrapped around AI and becomes part of what is shipped: capabilities that arrive already scoped, already evidenced, already auditable by the people relying on them, so trust is a property of the product rather than a promise about it. That is the stage HumanGrow is building towards.
Practice
For executives and boards: strategy, operating model, roles and accountability for AI. Clear decisions, written down, owned by named people.
Turning pilots into governed, measured capabilities. Product ownership and delivery discipline, from first prototype to production service.
The seven-step loop: register, trust level, scope, acceptance contract, quality ledger, decision log, promotion review. The Assay practice sits here, testing evidence before anything is promoted.
Practices under HumanGrow
Organisations
Roles held inside organisations, not a client roster. The detail sits in the story.
2026 –
Founder
An AI literacy collaborative, making practical AI capability accessible beyond specialist teams.
2026 –
Founder and CTO/CIO/CAIO
UK entity, company number 17386840. AI strategy, delivery and governance, with ENGRAM as the operating loop for every AI capability.
2019 –
Microsoft 365 Subject Matter Expert, then Modern Workplace M365 Product Owner
Microsoft 365 subject matter expert from June 2019 to April 2025, then Modern Workplace M365 product owner from April 2025. Platform capability across a global renewable energy business, then the move into governed AI.
2016 – 2019
Microsoft 365 Consultant
Client-facing consultancy delivery on a document management platform. Southampton and remote.
2012 – 2016
ITC SME Lead for Vodafone Partners
Led ICT capability development across the partner network, translating technology propositions into commercial outcomes for partner businesses.
2005 – 2012
Operations management
Operations management with full profit-and-loss ownership, compliance and service operations for high-volume retail sites, leading teams of up to thirty-three people.
Approach
Every AI capability is named and recorded before it runs.
A trust level, L1 to L5, sets what it may do. Never assumed.
An allow-list of the exact systems, data and actions permitted.
Executable acceptance criteria, agreed before any build.
A quality ledger, scored on reliability multiplied by utility.
The decision recorded, including the options rejected.
Promotion or demotion, on evidence only.
Nothing starts above L1 and nothing rises except on evidence.
The first version of this practice began before AI was part of any business plan. The belief was simple: organisations improve when the people inside them grow, and growth needs structure, honesty and patience.
That belief has not changed. What has changed is that AI is now one of the things doing the growing. The work is to keep human judgement in command of it, through clear roles, governed systems and evidence.
If AI is on your agenda and you want it governed properly, write or book a time directly.