Case studies

Selected work

Each of these follows the same structure — the problem, what investigation revealed, what was actually built, and what changed as a result. The reasoning matters more than the technology, so that is what gets the space.

Squava3+ years

Modernising a decade-old PHP platform without stopping the business

A ten-year-old PHP application was still earning revenue but had become slow and expensive to change. It was migrated to a modern architecture incrementally, in production, with API response times improving by up to 80%.

API performance improvement: Up to 80%API performance improvement
Continuous ownership: 3+ yearsContinuous ownership
Big-bang rewrites: ZeroBig-bang rewrites

Read the case study

PepsiCo1+ year

Turning scattered social and API data into something a business can read

A data-centric platform that pulls information from multiple external sources and social platforms, processes it, and presents it visually. Delivered end to end — backend, data pipelines and frontend.

External data sources unified: MultipleExternal data sources unified
Design through deployment: End to endDesign through deployment

Read the case study

Digital Healthcare Platform3+ years

Cutting video infrastructure costs on a platform built for doctors

A platform where doctors access video education, a social feed and peer consultation. Beyond backend delivery, a custom video transcoding service replaced a managed cloud service to bring running costs down.

Transcoder replacing managed service: In-houseTranscoder replacing managed service
Socket-based data updates: Real timeSocket-based data updates

Read the case study

PotatoAIHubLive and in active development

Building and running a live multi-model AI platform

A production AI platform bringing chat, image and video generation across multiple AI models into one interface. Built, deployed and operated end to end — proof that the AI work on offer here is practised, not theoretical.

In production at potatoaihub.com: LiveIn production at potatoaihub.com
Several AI providers, one interface: Multi-modelSeveral AI providers, one interface

Read the case study

Development EnvironmentsOngoing practice

Making “works on my machine” stop being a sentence anyone says

Standardising local development and QA environments with Docker, so that what a developer runs, what QA tests, and what production serves are genuinely the same thing.

To a running project: One commandTo a running project
Environment defined in the repo: VersionedEnvironment defined in the repo

Read the case study

Currently building

In development

Products being built right now. They are listed here because ongoing product work is the most current evidence of capability — but they have not shipped, so no results are claimed for them yet.

PotatoChat

Preparing production release

An Android application built around AI chat, with the production release currently being prepared. It covers the mobile side of the same problem PotatoAIHub solves on the web — a conversational AI interface with a backend that has to stay responsive on a phone.

Android
AI chat
Laravel API
AWS

Poise

In active development

A communication application being built in React Native, with a production release planned. It is the current piece of hands-on mobile product work — React Native, interface design, API architecture, and the path to shipping on a real app store.

React Native
REST APIs
Real-time messaging

Want the version of this for your own system?

The case studies above started as a conversation about something that was not working. Describe yours and you will get an honest read on what is actually going on and what fixing it would involve.

Discuss a Project Call me: 0490 916 967

Prefer not to call? Send me a message

Working with businesses across Australia — Melbourne, Sydney, Brisbane, Adelaide, Perth, Canberra and regional Australia.