statement
Full stack business from idea to profit. I build teams and strategies that convert investor goals and customer research into viable products. I’ve built new teams in hypergrowth (15 to 40 in six months), I’ve rebuilt teams and culture after restructures and have been trusted with products across health, agriculture, banking and payroll, including the one that pays one in seven Kiwis. Lately the work is agent-native: local models, agent context and infrastructure, and months of engineering work done in a day, repeatedly. One of them cost about $150 in tokens.
I’ve delivered cultural and technological change both ways: big bang greenfield builds and gradual, continuous improvement. I know when a business needs which.
I have found through practice that the happiness of your users rarely diverges from the happiness of your technical staff.
career
Smartly & Datapay (Datacom Group)
General Manager, Platforms & Architecture
HEADsmartlyI was promoted from Datapay into Smartly’s most senior technical role, reporting to the CEO. I am accountable for target architecture, engineering standards, platform-as-a-product and cross-entity architectural alignment with Datacom Group. Smartly is SME payroll. We pay 200,000+ people across 20,000+ companies. Smartly’s constraints are very different in comparison with Datapay: a smaller engineering team, pre-existing sprawl, less capital and a market that rewards simplicity.
Rebuilt the team: hired a platform team of eight back after the restructure that created my role, and set its ways of working: async-remote first, write things down, agent-native engineering.Made the platform’s state falsifiable: baselined the platform in my first quarter and published the headline measures the SLT now tracks, from architecture footprint consolidation to reliability and DORA measures. The business need can be met by something much smaller than the architecture we have today, so we measure the sprawl and reduce it, thereby reducing our cost to serve, while taking risk out of the business.Demonstrated what metric-driven investment looks like: Initially the budget I was set for modernisation was $1,000,000. I set the project up so that we drew down small chunks of that money to shift specific metrics. Where most might kick off a huge multi-year programme, I actively told my finance people “No! Don’t trust me! I don’t want the full budget. Let’s start with $20k for continuous delivery automation and see how that shifts our DORA metrics. Based on that, we can unlock more, or we can move on!”stackReact, React Native, .NET Framework 4 to 8, Microsoft SQL, T-SQL, SSIS, IdentityServer, Temporal, n8n, Fabric, Azure, Bicep, Datacom on-premPrincipal Architect
datapayDatapay is payroll software that pays one in seven Kiwis, over 250k payees across Australia and New Zealand. Datapay had the engineering team, the capital, and a board committed to a ten-year agentic payroll platform rebuild for all of Oceania. I was hired to ensure the technical success of that new platform. When I joined, my colleagues knew what the product needed to do but had no idea how to build it. An exhilarating challenge!
Designed and built Pacific, potentially the first agents-at-core payroll platform in Oceania: In 30 months, on time and on budget, validated in a parallel production run, the Pacific Platform ran seven customers’ live payrolls processed alongside the incumbent system with cross-checking. The pay packets matched for all the functionality we built for beta. We plotted “Percentage of pay packets correct to the cent” to create a view of “when are we done”. Pacific was shelved when Datacom Group’s investment priorities changed.Deployed an additional $1m of capital: The Datacom board asked if we could deploy an extra million in capital into our greenfields build in the last half of the financial year. We scaled product and engineering from 15 to 40, and I designed the organisation for fast flow so the new teams would produce the architecture we needed. Onboarding at that pace was disruptive. Fortunately I was able to retune the architecture and processes needed to support such a large number of new starters in that period of time. The existing team-topology-based approach I took to org structure and architecture was easily extensible: delivery quality and volume rose markedly.Got hands on to earn the mandate: Part of unlocking money early was demonstrating that we could reduce risk and make good use of the capital. I built 4 proofs of concept around specific pay and leave processes, encryption and security functionality, and even ended up building the complete fine-grained access control system for the platform. Those demonstrations helped stakeholders feel confident that the additional capital they deployed would be well spent.Delivered business value within six weeks of day one, via a rapidly deployed architecture process and principles. These principles fixed the log-jam of decisions teams were experiencing by ensuring three things happen for every decision we make: we seek feedback from those who are experts in the decision area, we seek feedback from those materially affected, and we write the decision and the feedback down. Almost overnight this converted capital A-Architecture from “something my boss or senior does” to “something I can do to learn and help the waka go faster”.stackLegacy: fat client, WCF, .NET Framework, custom ORM, Microsoft SQL, in-house identity provider Pacific: React, Flutter, .NET, CQRS, hexagonal architecture, event sourcing, KurrentDB, Postgres, Apache AGE, Marten, Auth0, Terraform, Azure, Datacom on-prem
FarmIQ
FarmIQ is farm management software. My business was strategy: developing, communicating and implementing both technology and business strategy.
Acting Head of Engineering + Lead Architect
farmiqWhile the Head of Engineering was on long leave, I acted in their stead. I managed ~ 10 engineers. During that time I built growth plans, managed the technology estate and ran training for the engineers, mentored juniors and sponsored intermediates and seniors so that they could create impact in areas they felt passionate about.
Lead Architect
farmiqWhen I started at FarmIQ, it became clear that as a business we were reactive to opportunity rather than working to a plan. My biggest job was not to cargo-cult in a strategy nobody wanted, but to help key leaders in the business understand where strategy could help them with challenges that didn’t suit a reactive approach.
Created the first holistic business and technology strategy FarmIQ has ever had: In the open and with leadership buy-in, I ran a collaborative process that built on the product vision. I pulled in data from market, from customers, from staff and from leaders, rapidly assembling a business and technology strategy that right-sized the product and technology investments against the market opportunity. Having that strategy helped staff on the ground connect the activity they were carrying out with results, which in turn caused the team to naturally plan for impact, using data to validate that impact. Previously they had built great product but the investment could not be connected with the impact or results the investment created.Reshaped software teams around cognitive load: Created a light, team-topology-focused structure that got engineers understanding the domain better with less collision between teams trying to work on different features in the same codebase.stackReact, React Native, Java, Postgres, Microsoft SQL, SSIS, Auth0, AWS
Livestock Improvement Corporation
Three out of four NZ dairy cows are sired by an LIC bull. In 1992/93 (first season recorded) the average amount of milksolids produced per cow was 259kg. By the 2024/25 season, that number increased by almost 60% to 414kg. NZ Animal Evaluation attributes around two-thirds of the gain since 2005 to genetics.
Platform Architect
licI was promoted into a newly created role as the first hire of the organisational structure I designed. The Platform Architect role is a fusion of product management and architecture. The goal of the role is to create an underlying platform that our customers use every day, but will never see, while remaining close to our engineering and end user customers to make sound roadmap and investment decisions. The real trick is that for someone with the right experience, the investment choices we make naturally result in good system design choices, because the person making the investment and roadmap decisions is the same person who is ensuring the architecture and engineering meet the financial reality without making poor trade-offs.
Launched a first-of-its-kind-for-NZ smart animal wearables integration: During my time at LIC a fastbal1l project came up: we wanted to beat the competition to the first meaningful smart animal wearable integration with MINDA, our on-farm herd management software. In 10 weeks I put together a team of ~ 4 engineers plus one sales and marketing colleague. We used one-day sprints to build the software and went live two weeks ahead of schedule, returning the initial investment plus profit two weeks before launch. I won an industry award for that work! The code was simple, but high quality. A single deployed API and a database, with a significant investment into non-functionals like observability, reconciliation, billing and security. My main error was not documenting why we made those tradeoffs: the engineer who inherited the codebase could reverse-engineer the how, but never the why. My fault, not his.Re-architected a critical animal genetic analysis system: Animal Evaluation (Herd-by-Herd) is the beating heart of the outsized impact on GDP that LIC produces each season. It was a collection of bash scripts and C files glued together with Python code. Essentially, it was 26 million simultaneous equations that would run on every single EC2 instance AWS had in Sydney above a certain size class (they had to buy more because of us). I redesigned that system such that it became serverless, event-driven and real-time.Solution Architect
licI was hired into the strategy and architecture team to consolidate the stack while moving the older systems into modern architectural paradigms. LIC maintained three epochs of architecture across 20+ production systems covering dairy herd management, genetics platform and first-mover IoT integrations.
Stayed in the code: I ran peer and mob programming, whiteboarding designs, algorithms and API contracts with the teams. I worked mostly on MINDA, LIC’s herd management software, designing and helping to build improvements to animal medicine management, health information and multi-herd functionality.Introduced the product owner role and software product management thinking into the technology group.stackReact, React Native, .NET Framework, Python, Bash, event-driven architecture, scientific computing, Azure, AWS
Bank of New Zealand
Digital Experience Product Owner
bnzMy product was bnz.co.nz. I left BNZ because the behaviours and priorities expected of me by the bank were then and are now incompatible with my own beliefs and values.
Healthlink
Product Manager
healthlinkI was promoted from engineering, and took on the challenge of moving Healthlink from a forms software provider to an ecosystem provider.
From seven words to a product line: the combined knowledge the business had around what they wanted me to deliver was seven words. The first cab off the rank, a provider directory with search integrations, launched six months later: six enterprise customers covering ~ 80% of the Australian market and 100% of the New Zealand market, ~ 500k in project costs over 4 months, subscription revenue model post launch.Software Engineer
healthlinkHMS is the software that your e.g. radiologist uses to collaborate with your GP on that nasty broken arm you got while wind surfing.
Implemented Continuous Delivery, with a colleague, later adopted business-wide: worst-case deployment went from 3+ months to 2 minutes; 60+ automated deployments in a busy week, tooling upgrades reaching 40+ developers and testers.stackAngular, Java, Elasticsearch, Solr, Lucene, Microsoft SQL
Spark Ventures
Junior Software Engineer
sparkI started as an intern; converted to full time within months. I shipped Morepork (IoT home security: too early to market, in my opinion), Bigpipe (the ISP for techies and gamers; proud to remain a customer) and Lightbox (a homegrown Netflix competitor, sold to Sky TV).
stackAngular, Java, AWS
NZ Army
Intelligence Operator (Returned), NZ Intelligence Corps
nz-armyNine months in Afghanistan on Operation Ariki, across TU Crib 18 and 19. Postings to 1 NZ Military Intelligence Company, 1 NZ SAS Regiment and 3 Signal Squadron (Electronic Warfare). Operational service medals. Self-taught networking and systems administration, then built a sandboxed network of computers that formed a deployable secured GIS cell. I had to learn things quickly. There was no training.
industry contributions
- recognition
- Emerging ICT Leader of the Year, CIO Awards, 2023.
- writing and speaking
- The Helpful Web (2025) · Architecture is gardening (2024) · A framework for creating strategy (2022) · The Role of a Software Developer (2022) · Implementing a team topology (2021).