Story
Twenty-three years of growing people. Now AI grows too.
Leonard Hrihoreanu is an AI leader working at CTO and CAIO level, and the author of the ENGRAM AI governance framework. The route here ran through operations, enterprise delivery, consultancy and platform leadership, in that order, and it matters.
Each period taught one thing that the next period needed. Governance was learned on forecourts before it was written down for artificial intelligence.
The timeline
2003
Human Grow
Founder
The practice began as Human Grow: training, coaching and the unglamorous work of helping people become better at their jobs. The premise was that organisations do not improve by decree; they improve when the individuals inside them grow, and when that growth is structured rather than hoped for.
It was a small operation, but it set the method that still applies: agree what better looks like, do the work, then measure honestly whether anything changed.
Growth needs a definition before it needs a programme.
2005–2012
Fuel retail
General Manager & Partner
Seven years as General Manager and partner in fuel retail. A sector where safety, compliance and margin are all measured daily and none of them forgive improvisation.
This was the operational apprenticeship: rotas, standards, audits, incidents, and the slow business of raising performance across people who did not choose you as their manager.
Process without people fails; people without process burn out.
2012–2016
Vodafone
Senior ICT, Microsoft 365 Expert
As a senior ICT and Microsoft 365 expert inside a large enterprise, the lesson was how change actually travels: governance forums, service standards, and dependencies between teams that never meet.
The lesson was about scale. A good idea delivered without owners, evidence and a route through governance does not survive contact with the organisation.
At scale, the hard part is not the idea. It is the accountability around it.
2016–2019
Microsoft 365 consultancy
Microsoft 365 Subject Matter Expert
Cloud adoption and modern work programmes for client organisations: migrations, collaboration platforms, security and identity, and the change management around all of it.
Working across many organisations in a short period made the pattern obvious. The technical migration was rarely the risk. Whether anyone changed how they worked afterwards was the risk.
A platform is only adopted when someone is accountable for the behaviour change.
2019–
RES Group
Product Owner of the Modern Workplace, then AI leadership
Leading Microsoft 365 and modern work first, then artificial intelligence: platform strategy, policy, delivery and the operating model that has to sit underneath both.
This is where AI stopped being a demonstration and became an estate: many use cases, many owners, real data, real consequences. It also became clear that the standard tools of enterprise governance were not built for systems that produce different output every time they run.
AI needs governance that measures behaviour over time, not documents signed once.
2024
The ENGRAM framework
Author
ENGRAM was written from deployments rather than principles: a seven-step loop of register, trust level, scope, acceptance contract, quality ledger, decision log and promotion review, with a trust ladder from L1 to L5.
Its single rule is the part that changes behaviour. Nothing starts above L1 and nothing rises except on evidence. Which means autonomy is earned in a review, with a ledger, in front of the people accountable for it.
Define the test before you see the result, or success gets redefined afterwards.
2026
Fidevo, AI Labs and HumanGrow
CAIO at Fidevo; founder of HumanGrow
New ventures in applied AI, and the original practice brought back under one word: HumanGrow. Advisory for executives and boards, delivery management for teams turning pilots into governed capabilities, and governance work built on ENGRAM.
The name still means what it meant in 2003. The difference is that AI is now one of the things doing the growing, and human judgement has to stay in command of it.
Human first. AI native.
How the work is done now
The practice is deliberately small and senior. Engagements begin with what already exists rather than a strategy document: what AI is in use, who owns it, and what evidence there is that it works.
From there the work is registering, levelling and reviewing, until the organisation can run the loop without help. The measure of success is that the framework outlasts the engagement.
If that way of working suits your organisation
The door is open. Write directly, or book a conversation.