About Graymatter
Graymatter is led by Phil Gray, an enterprise innovation and change practitioner who has spent more than three decades helping organisations move from an idea to a different way of working.
Today that work is AI: choosing the right use case, building the proof, earning adoption and keeping a human clearly accountable.
Phil Gray · Founder
Phil leads the discovery, shapes the engagement and stays involved through the build, tests and decisions. He works directly with the people who own the outcome and the people who understand the workflow.
Graymatter is deliberately small. The model is senior, hands-on delivery supported by a governed AI production system, with no gap between the person who frames the work and the person accountable for it.
Email PhilThe useful part of the CV
AI adoption is a technology problem, an operating-model problem and a human change problem at the same time. Phil’s background crosses all three.
Phil built and led an internal innovation accelerator inside a major Australian bank. Its job was to help a large, regulated organisation run smaller, faster experiments without losing sight of the customer or the path to scale.
As global head of a Singapore-based innovation lab for a major international bank, Phil led fintech experiments across areas including AI, blockchain and regulatory technology, as well as the harder work of moving useful ideas towards operations.
Across business, HR, change and transformation, Phil has worked on what happens after the technology decision: ownership, behaviour, leadership, capability and benefits.
Phil now designs and operates bounded agent systems for research, evidence handling, workflow coordination, human review, monitoring and executive reporting.
See project patterns ↗Phil has also spoken to CIO and executive audiences internationally about disruption, experimentation and why innovation must make the leap from a lab into live work.
Inside our delivery model
AI agents accelerate bounded parts of Graymatter’s own work. Phil remains accountable for judgement, client context, quality and action.
Client objective, risk appetite, commercial judgement and final approval.
Research, compare, prototype, test, structure evidence and prepare documentation.
Workflow knowledge, source information, users, constraints and adoption decisions.
Working evidence with a person clearly responsible for what leaves the room.
Working principles
A working first version teaches more than weeks of private planning. You see the brief, choices, prototype and test evidence as they develop.
The right answer may be a simpler automation, a source-data fix, a different workflow—or no build yet.
Sources, uncertainty, recommendations and unresolved issues are labelled when the distinction matters.
AI can prepare and recommend. Accountable people own consequential business, client and workforce decisions.
Briefs, tests, records, working guides and decisions should remain useful if tools or delivery partners change.
A prototype, scaffold or successful build is not proof of live performance. Claims follow the evidence available.
A wider team when needed
Some work needs deeper capability in cybersecurity, privacy, legal, data engineering, enterprise platforms or organisational change. We identify that need early and can work alongside your existing advisers or agreed specialists.
The Graymatter role remains clear: keep the pilot connected to the business outcome, integrate the work and preserve one decision record.
A practical first conversation
Start with the workflow, the reason it matters and what has stopped it improving so far. You do not need to arrive with an AI specification.
Start with the workflow