I build products end to end: the frontend, the backend, the infrastructure, and everything in between. Give me a problem and I'll own it from the first commit to production and keep it running after that.
- 🧩 I'll take on any problem. AI agents, backends, frontends, infrastructure: if the product needs it, I build it.
- 🎨 I'm a product person. I care a lot about UI/UX and I notice problems in things that look fine to everyone else. I keep refining an experience until it feels right, because good products come from good experiences.
- 🚀 I take ownership from zero to shipped. That means design, code, tests, deploys, observability, and support.
- 🤖 I'm AI-forward. I build AI products, I use AI coding agents every day, and I follow everything happening in the field.
📍 Melbourne, Australia · If I'm not shipping something, I'm probably benchmarking a new model or dying to a boss in Elden Ring.
- AI agent platforms. Agent runtimes that run durably, stream results in real time, and connect to tools, memory, and MCP. I build them for the web and for the apps people already work in.
- Full-stack products. Internal platforms covering analytics, search, content management, and admin, built across both the frontend and the backend.
- The infrastructure under them. Auth, queues, observability, CI/CD, and cloud deployments, all built and run by me.
I usually take a product from an idea to production on my own, then keep improving it.
I have a machine at home for running models locally, and it's where I spend a lot of my free time.
- I serve open-weight models with vLLM, SGLang, and Ollama, and compare how they perform on my own hardware.
- I build RAG pipelines and run bulk data-processing jobs to see what actually holds up outside a demo.
- I try new model releases and architectures soon after they come out, then decide what's worth bringing into my own work.
I follow the industry closely. I take it seriously, but mostly it's fun: I like learning new things and solving problems.
I'm mostly self-taught. I did get my degrees, but I learned the most by building things.
In college I spent most of my time in Blender, Unity, and Unreal Engine, making 3D games and environments. My biggest project was a recreation of PUBG's Playground mode:
- I modelled every asset by hand.
- It took about two months to build the full realistic environment.
- Along the way I learned the whole game pipeline: modelling, texturing in Substance Painter, sound design, composition, camera work, animation, and scripting.
Blender is still my go-to tool, and I'm always happy to open it again.
These days I spend my free time in Elden Ring and Dark Souls, or playing Valorant and Apex Legends. Souls games taught me a lot about debugging: die, learn the pattern, try again.
Most of what I build lives in private repositories at work, so it doesn't show up on GitHub by default.
My contribution graph includes that activity as timestamp-only commits in a private repo. No code, commit messages, or project names leave the original repositories.
AI: Vercel AI SDK · OpenAI / Azure OpenAI · MCP · agents & tool calling · RAG · evals · AI coding agents
Local AI: vLLM · SGLang · Ollama · open-weight models · RAG pipelines
Frontend: TypeScript · React · Vite · Tailwind · TipTap · D3 · Office.js
Backend: Bun · Hono · Express · Convex · Redis / BullMQ · MongoDB · PostgreSQL · Elasticsearch
Infra: Azure (Container Apps, Static Web Apps, Blob, Entra ID, App Insights) · Docker · Envoy · Nginx · OpenTelemetry · Azure Pipelines
3D & games: Blender · Unity · Unreal Engine · Substance Painter
