Profile
Background
I’m a senior software engineer with more than ten years of experience. I work mainly in
TypeScript/Node.js, C++ and Java, and predominantly in cloud environments on AWS and GCP,
with Docker, serverless architecture and AI/ML integration as everyday tools.
Most recently I architected an LLM-enabled agentic migration pipeline at Squiz. Before
that I led infrastructure load testing and optimisation on GCP at Safewill, and
spearheaded the architecture of an AWS SaaS platform at Comet Strategy that supported
more than 1,000 container instances.
I’ve worked both within agile teams and independently as the key architect on a product.
What I do best is solve complex problems: operationalising AI, automating processes, and
delivering robust, high-value software systems.
The company
Vierte Pty Ltd
Vierte is the company I founded in 2022 for independent work: contract engineering,
architecture reviews and scoped software projects. When you engage Vierte, you work
with me directly.
I’m open to permanent senior engineering roles as well as contract work.
Projects
Selected work
Agentic migration pipeline and LLM observability
Squiz · 2025–2026
I architected and built a scalable migration platform that uses large language models to convert legacy websites onto Squiz’s platform. The models generate the components, assemble full pages from the legacy content, and then review and refine the result.
Alongside it I delivered a cost-metrics framework for the AI team, so the business could see token consumption and cost per model and per agent. I also modelled input caching for function-calling agents, which became the blueprint for reducing token spend on large migrations, and developed “dehydration” algorithms that shrink LLM context windows to reduce latency and cost.
Comet cloud service
Comet Strategy · 2017–2021
I was put in charge of moving the company’s product, Comet, to the cloud, and did the architecture and the full development myself. The result was an AWS SaaS platform for massively parallel workloads, running 1,000+ container instances per user.
Getting there meant significant groundwork first: porting the Windows application to Linux, adding automated testing, and building a continuous-delivery pipeline. The service went on to become a paid product for the company’s clients.
Load testing, data privacy and a platform move
Safewill · 2022–2024
I led GCP load testing and infrastructure optimisation so the platform could scale cleanly, and led the solution architecture for a data-privacy solution that tokenises personally identifiable information. I also moved the stack to NestJS to improve consistency and development speed, and ran technical workshops to bring the teams along with it.
Experience
Career history
May 2025 – Jul 2026
Senior Software Engineer, AI Team · Squiz
Built an agentic migration pipeline on AWS Bedrock, Lambda and S3 to automate moving legacy sites to modern architectures. Researched and modelled input caching for function-calling agents, an optimisation adopted to reduce token usage on large LLM migrations.
Jun 2022 – May 2024
Senior Software Engineer · Safewill
Led GCP load testing and infrastructure optimisation, and the solution architecture for tokenising PII data. Transitioned the tech stack to NestJS and upskilled teams through technical workshops.
Jul 2017 – Sep 2021
Senior Software Engineer · Comet Strategy
Designed, architected and developed an AWS cloud platform for massive parallelisation (1,000 machines per user), after porting the Windows application to Linux and building automated testing and a continuous-delivery pipeline. Also proposed and explored new algorithms, products and business ideas.
Oct 2014 – Jun 2017
Software Engineer · Expedia and Wotif
Worked mainly on Java backend services with the Spring framework, deployed on AWS, including one of the search and booking platforms and a developer API for hotel market insights. Built the front end and Node.js backend for the developer API website.
Feb 2014 – Oct 2014
Software Tester and Developer · Suncorp
Tested Java applications end to end on Linux servers, and automated tests and deployment. Primary tester for the health insurance product sold under the NIB brand.
Skills
Languages and tools
Education
Bachelor of Engineering (Software)
The University of Queensland · 2010–2013
My undergraduate thesis used Deep Belief Network (DBN) machine-learning techniques to
segment cervical cell nuclei.
