Product engineer · Auckland, New Zealand

Spencer Hu

I build production software end to end: workflow systems, background processing, public APIs and the interfaces on top of them. Nine-plus years in, recent work includes applied AI, both in the products I build and in how I build them.

experience

Professional work

Arbitr

Formerly Straker Translations · rebranded 2026

  • Product Engineer2026 — now
  • Intermediate Software Engineer2021 — 2026
  • Software Engineer2019 — 2021

Arbitr provides translation services and AI context management. I work on the platform behind it: the workflow that moves a job from upload to delivery, the background processing underneath, the public API around it, and the AI features built into it. This is team work; each item below states my part.

Led much of the architecture

Core translation workflow

The system that coordinates a translation job: file processing, vendor selection, the steps where people are involved, and the conditional business rules that decide what happens next.

Designed

Background processing

Redis-backed asynchronous jobs for machine translation and file processing, keeping slow work off the request path. It has run thousands of tasks.

Built

Public API

An API for external integrations, with API-key authentication, so clients can create and manage translation projects programmatically and receive webhook notifications.

Built and contributed

Applied AI

AI-assisted email classification and content categorisation, vendor ranking, and terminology/glossary matching inside the translation workflow.

Day to day

Product engineering

Working closely with Product to shape features, then developing, reviewing and delivering them, using AI coding tools throughout.

Earlier: Java Web Developer · Inspur Group, 2016 — 2018

projects

Built independently, in my own time, to explore ideas end to end. These are personal projects, not commercial products.

about

I like owning a feature from the first conversation to production: understanding the problem with Product, designing the data and the workflow, building the backend and the interface, and staying around to see how it's used.

Most of my experience is backend and integration work: workflow engines, background jobs, public APIs and the business rules in between. I still care about the interface on top, because that's where people decide whether the system is any good.

A lot of my work has been modernising production systems that people rely on every day, so I tend to be pragmatic: small, safe steps, honest trade-offs, and nothing that breaks what already works.

I studied Computer Science at the University of Waikato and Communication Engineering at Northeastern University, and I speak English and Mandarin.

Skills

Backend and architecture

Python, Django, FastAPI, Redis, WebSockets, Elasticsearch, Java

Frontend and UI

TypeScript, Vue, Nuxt, React, Tailwind CSS

Data and infrastructure

PostgreSQL, MySQL, Docker, NGINX, GitHub Actions, Fly.io

Applied AI and tooling

OpenAI API integrations, structured outputs, retrieval (RAG), AI coding agents