Processes span multiple systems (Project Planning, Scheduling, Demand Forecasting, Logistics and more), often stitched together manually or reliant on brittle workarounds. Offline spreadsheets, siloed data, and legacy platforms result in limited visibility, duplication of effort, and constant context-switching. There's no single source of truth — and no way to see, let alone optimise, the whole picture.
Starting from broken trust.
I joined after a failed MVP launch that users had flatly rejected. Morale was low and trust — from users and stakeholders alike — had eroded.
The MVP had been shaped around SME assumptions rather than user insight. It digitised a narrow ordering process without accommodating the real-world complexity around it. It didn’t add value — it added friction. And it wasn’t scalable beyond oil & gas.
Bringing clarity to complexity.
I led product and design direction from the ground up — building the team, shaping strategy, and defining our AI vision. My responsibilities included:
Led product & design strategy
Hired and managed a team of Product Designers
Introduced service blueprinting, systems thinking, and mixed-method research to shape direction
Defined the AI strategy and prototyped agent-powered experiences — from opportunity mapping to generative and agentic UX concepts
First principles. Fast feedback.
We stepped back to understand the full system — deeply researching archetypes, mapping a sprawling service blueprint, and identifying where value was leaking.
We delivered modular, scalable solutions in fast feedback loops with users at the centre, gradually rebuilding credibility and momentum around a clear future vision.
●
WORKFLOW ENGINE
Modular automation
for messy processes.
A flexible system of triggers and actions to orchestrate real-world workflows across fragmented systems — built to adapt, not constrain.
●
AI Agents
Intelligent
operational copilots.
AI agents replaced manual effort in key workflows while also acting as on-demand copilots to assist users with tasks.
●
KNOWLEDGE HUBS
Signal from the noise.
Operational data is transformed into a queryable layer — allowing agents and humans to retrieve answers instantly, without relying on tribal knowledge or offline workarounds.
●
KEY BUSINESS OUTCOMES
Reduction in material waste
40%
(~0.25% of global CO₂e emissions)
Reduction in manual hours through automation
64%
time savings