AI systems that can operate
- Agent and MCP architecture
- RAG and grounded knowledge
- Evaluation and structured outputs
- Permissions, audit and human hand-off
Brisbane · Available for the right engagement
AI is moving fast.
Useful systems
still have to ship.
I turn ambiguous AI and software opportunities into working products, safer automation and faster decisions—drawing on more than 20 years of hands-on delivery.
20+years shipping software
15+years mobile & cloud
AIagents, automation & products
01 / What I do
I’m most valuable where the problem matters, the path is unclear and a slide deck is not enough.
I can move from targeted research and opportunity mapping through architecture, hands-on implementation, integration, testing, release and team enablement. The goal is not to “add AI”. It is to improve how the organisation works.
02 / Capability
One operator.
Three kinds of leverage.
Software that survives reality
- Mobile and native products
- Cloud APIs and integrations
- Offline-first and recovery design
- Testing, CI/CD and release
Decisions that improve the business
- Opportunity and workflow mapping
- Product and commercial strategy
- Targeted deep research
- Build, buy, automate—or stop
03 / Selected evidence
Built across the stack.
Judged by the outcome.
Current AI systems, native products, enterprise delivery and decision work—labelled by their real state.
01Private beta
Browser intelligence
PageSift
Turns rendered web pages into compact, structured context for AI tools—locally, with deterministic extraction beneath optional on-device intelligence.
- Chrome MV3
- TypeScript
- On-device AI
- Privacy
02Public
Agent infrastructure
MCP Conductor
A sandboxed MCP orchestration layer that lets AI clients compose tool work while returning only the result that matters.
- MCP
- TypeScript
- Deno
- Sandboxing
03In development
Native macOS systems
Lumnit
A coherent display command centre spanning hardware DDC, built-in panels, XDR, scenes, automation and failure recovery.
- Swift 6
- SwiftUI
- Hardware integration
- Concurrency
04Client delivery
Accessible product modernisation
BathCheck
Long-lived mobile and web product work supporting safer bathroom-access decisions, modernised across .NET, MAUI, web and cloud.
- .NET MAUI
- Blazor
- Mobile
- Cloud APIs
05Historical delivery
Government · transport · health
Enterprise mobile systems
Hands-on architecture and delivery across the ATO mobile app, V/Line staff and passenger systems, Telstra HealthNow, Queensland Sugar, Maxxia and RemServ.
- C#
- Xamarin
- .NET MAUI
- Azure
06Client delivery
Research to operating action
Decision intelligence
Targeted research that connects market context, customer behaviour, campaign choices, workflow design and practical automation.
- Deep research
- Opportunity mapping
- Automation
- Commercial systems
04 / AI-assisted, human-accountable
I use AI as leverage—not as authorship theatre.
I own the problem framing, product judgement, architecture and trade-offs.
AI agents help research, explore, implement, test and document at greater speed.
Evidence, review, safety boundaries and the final result remain my responsibility.
05 / Ways to work together
Embedded
Put me inside the hard problem.
High-value 3–6 month contracts, technical leadership and delivery recovery.
↗Dark IceBring a focused outcome to life.
Consulting, project delivery and ongoing architecture through the studio.
↗Leadership roleBuild the capability for the long term.
Selective principal, architecture and technical-lead opportunities.
↗