Case studies

What I led, and what it delivered.

Three roles, told the same way: the problem, what I did as the leader, the team I led, and the result.

Enterprise AI2025 to 2026

Taking AI from early ideas to products in real use.

Director, Data & AI · Group-wide AI and data leadership

The problem
The group had no shortage of AI ideas across its businesses, but no shared strategy for which to back and no common standard for taking them safely into production.
What I did
I owned AI and data strategy across the group and ran it lean: one prioritised backlog, with each idea tested as a minimum viable product before we committed more. I set the governance and lifecycle standards every release had to meet, and directed the agentic AI work in thin vertical slices, so every increment worked end to end.
The team
A cross-functional, self-organising team spanning machine learning, computer vision and full-stack engineering. I set the structure, a clear definition of done, and a short sprint and review cadence that kept feedback loops tight and work shipping.
The result
AI moved from early ideas to products in real use through small, frequent releases, built to one governance standard rather than a scatter of one-off pilots.
Philanthropy2020 to 2025

Evidence, not instinct, behind investment and policy.

Manager & Principal, Data Science · Minderoo Foundation

The problem
One of Australia's largest philanthropic foundations was investing across many programs at once. It needed evidence, not instinct, to decide where money and policy effort would make the most difference.
What I did
I led data science strategy for the foundation. I built and ran a foundation-wide data roadmap and capability program, and embedded machine learning and predictive modelling into core programs so analysis fed real decisions.
The team
The data science team, which I managed, working side by side with program leads so models answered the questions decision makers were actually asking.
The result
Predictive models and analytics that guided investment and policy decisions, and a stronger data capability across the foundation.
National platforms2019 to 2020

Building and mentoring a team from the ground up.

Data Science Manager · Elm, Riyadh

The problem
National digital platforms in Saudi Arabia needed machine learning models for real-time, intelligence-led decisions, drawn from mixed and messy data. The team to build them did not yet exist.
What I did
I built the team from the first hire, set the technical and working standards, and mentored each member as they grew into the work. I kept the models tied to the decisions they had to support.
The team
An all-female machine learning team: talented specialists who became a high-trust unit that delivered to a national standard.
The result
Models delivered into national platforms, and a team that proved capability has nothing to do with gender.
Read: Knowledge knows no gender

Products

Strategy that ends in shipped product.

AI-assisted Data Intelligence Products

A portfolio shaped and shipped under my direction, proof that the strategy ends in something real.

Domain depth

Mining, healthcare & safety-critical AI

The sectors I know, where I set AI and data direction when the data is messy, the stakes are physical and the margin for error is small. From mining and resources to hospital triage, ambulance dispatch and road safety.

MiningResourcesIndustrialMedical triageAmbulanceRoad safetyHigh-risk decisions

Looking for an AI leader in Perth?

I'm open to Head of AI and AI strategy roles. Let's talk.

Get in touchDownload CV