Agentic AI Governance
Oversight for semi-autonomous systems inside regulated guardrails.
Advisory practice
AI Governance & Change Architecture for high-stakes production systems.
I help complex organizations govern how AI agents, digital products, and production-impacting change reach deployment: safely, proportionately, and with defensible go/no-go decisions.
Where governance has to hold
AI agents do not only fail in demos. They fail where delivery, controls, ownership, and go/no-go decisions meet.
Domain pillars
Where production reality meets executive judgment.
The work is not policy theater. It is the operating layer that decides what reaches production, under which controls, with which risk trade-offs, and why.
PDLC & Go/No-Go
The decision layer for safe, reversible, accountable releases.
Sensemaking for Change
Frames for non-linear risk and technical uncertainty.
Scale15+
years across technology, financial services, industrial, energy, and mobility environments
Scope1,500+
Agile teams in transformation and governance context
Impact85M
clients impacted through regulated banking production systems
Judgment100+
production-impacting change intakes assessed through go/no-go logic
Operating logic
A decision trail, not a slide deck.
1
Map the ambiguity: who owns the pipeline, who sets controls, who carries production risk.
2
Calibrate the control logic: what is justified, proportionate, measurable, and reversible.
3
Translate the result into executive language: risk, ROI, governance, adoption, and decision rights.
Ambiguity
Risk logic
Governance
Go/No-Go
Field notes
Frameworks for the AI production edge.
Short, executive-readable notes on agentic AI governance, decision gates, control-induced risk, and the systems that make high-stakes change legible.
Governing AI agents in regulated PDLC
How agentic systems change the release boundary, not only the technology stack.
Open notesThe independent go/no-go layer
Why the strongest governance challenges both delivery owners and control setters.
Open notesWhen controls create second-order risk
A practical frame for spotting when safety mechanisms become production risk.
Open notesMore than one discipline
Beyond the advisory work.
Governance is the spine, not the whole picture — teaching, painting, and writing train the same way of seeing. Read the through-line.
Teaching & speaking
London School of Mathematics & Programming, conference talks, and cross-institution mentoring.
TeachingStudio
Oil painting, exhibitions, and writing — structure, proportion, and hidden tension as a practice.
StudioAbout
How engineering, governance, teaching, and art are one habit of seeing.
AboutSelected recognition
Independently endorsed.
UK Global Talent — Digital Transformation · Cambridge MBA, Chevening finalist (top 1%) · MIT, AI in Business Strategy. Full recognition and education.