Work
More than 17 years in software, the last 7 of them leading teams. Roles below run newest first. The résumé has the compressed version, and projects covers what I built rather than where I built it.
Engineering Manager and Technical Lead
Big Viking Games, Toronto · April 2025 to present
Founding engineer on the ground-up re-platform of YoWorld, a social MMO live since 2008, from Flash onto a Unity 6 WebGL client running in the browser on desktop and mobile.
I wrote the initial commit and set technical direction from there. I own the architecture across three surfaces that have to agree with each other: the Unity client, the realtime multiplayer server, and a serverless cloud backend. I also built the 14-person team behind the rebuild, spanning frontend, backend, art, and QA, and hired and onboarded every one of them.
Nine of that team report to me directly: five senior Unity engineers, two senior backend engineers, a lead QA, and a graphics programmer. Four of them had been leads at other companies before joining. Six more people across art, DevOps, and product do not report in, but I plan and direct their work. Day to day it runs as three sub-teams: art, backend, and the Unity realtime-server and frontend group.
Some of the job reaches past my own reporting line. I direct and advise other engineering managers on AI enablement, and I take part in cross-team architectural review. For four months I covered the CTO's responsibilities, during the vacancy between the outgoing CTO and their replacement. I advise the CEO and COO directly on company technical direction and take part in company-level quarterly planning. Scope and resource estimation for the rebuild is mine too: headcount allocation across the sub-teams, software and backend service cost estimates, and the return-on-investment cases behind them.
The project reached its first external beta in July 2026. Players scored it 4.6/5 on excitement to return the next session.
Over the same period I led AI enablement for the company's engineering department. In four months it went from no AI usage to 52% of merged pull requests being AI-authored or AI-assisted, and 8 developers now ship production code through agentic tools. Every AI pull request clears the same continuous integration and human code review as hand-written work.
People development runs on a schedule rather than when time allows. Every report has a monthly one-to-one carrying two personal growth goals, and I match those goals against work the business needs done, so the development time pays for itself. Each quarter ends on a real deliverable, an internal or external demo, followed by a gathering to mark it.
I set the security posture for the platform: server-authoritative design for everything critical, authenticated and rate-limited backend calls, and no anonymous access to backend services. I drove a sustained migration from client trust to server authority, and directed hardening work on identity, secrets, and the login path. I also built a scheduled weekly audit that sweeps the backend and reports on its own, which makes security review routine instead of reactive.
The rest of the job is the unglamorous half: coding standards, agile process, CI/CD, automated testing, and infrastructure-as-code.
The old product is still live, and keeping it that way is part of the work. I contribute to the maintenance and refactoring of a 15+ year old codebase while we build its replacement, and I direct the interop services bridging the two. I also supervised delivery of a documented, repeatable pipeline that converts legacy Flash content into modern engine assets.
Parental leave
Toronto · March 2024 to April 2025
Senior Unity Developer
Forma Vision, Seattle (remote) · April 2023 to March 2024
Lead developer on the real-time 3D meeting-room clients that rendered people as volumetric holograms, used by Accenture, Fidelity, Stanford Medicine, and Cisco. The role ended when the company wound down.
My main contribution was a multi-platform delivery model: a feature was built once and shipped to desktop, mobile, and six XR platforms, namely Meta, Magic Leap 2, Apple Vision Pro, Snapdragon Spaces, XReal, and OpenXR. That model is what kept roughly 8 distinct client applications sustainable across desktop, mobile, and standalone headsets. I architected and owned several of the shared internal packages underneath it, and advised other teams on adopting them and contributing back. One of those was a reusable UI package, so AR, VR, and desktop could share components instead of forking them.
I also wrote the 3D transformation and calibration systems that aligned several synchronized Intel RealSense depth cameras to a virtual scene. That was a live capture pipeline, calibrated on-site in customer studios.
Real-time rendering and GPU performance was a constant across the whole job. Volumetric hologram playback has to hold frame rate on standalone XR headsets as well as desktop, and that shaped a lot of the 3D pipeline work.
Elsewhere in the role I led refactoring work for decoupling and testability, ran customer demos of holographic collaboration, and mentored junior developers on software architecture.
Lead Developer
Wizards of the Coast (Hasbro), Toronto · November 2021 to April 2023
Lead developer on a Peppa Pig mobile educational app. The project was later sold to Netflix Games.
The role covered more than that one app. There was also a Play-Doh branded product I contributed to and advised on strategy for, and the company-wide platform product that every game was built on. The platform is the half that mattered most, because it was shared-foundation engineering rather than another title on the shelf.
I managed the developers and owned every software architecture decision from inception through alpha release. The localization work went deeper than translated strings: I architected the remote content pipeline on Unity Addressables, Localization, and Cloud Content Delivery, so new content and new languages could reach players without shipping a client update. Alongside it I built a custom localized voice-over system on Unity Timeline and Playables, which the company adopted as its standard animation workflow.
I built reusable internal libraries for animation, character rigging, and scene management that were shared across products, and wrote the technical design documents that kept artists and developers working from the same picture. I also made the case to management for long-term investment in code infrastructure, which is where test-driven development, automated testing, and CI/CD from pull request to store build came from.
Mentorship was a formal part of this one. Career planning with team members, and monthly seminars on architecture and best practices.
Lead Developer and Research Associate
Centre for Vision Research, York University, Toronto · September 2018 to November 2021
Lead developer at York's Centre for Vision Research, building immersive VR and AR tools for studying human perception.
Two open-source Unity SDKs I shipped there, bmlTUX and bmlSUP, were picked up by vision-science labs internationally. I still maintain both. bmlTUX was peer-reviewed in i-Perception.
On the systems side I built real-time body-matched self-avatars driven by OptiTrack motion capture, the tracking-to-rendering pipeline behind them, and multiplayer server fleets on AWS running authoritative Unity DOTS NetCode. A separate project was a multi-user AR app built on ARKit face and body tracking with OpenCV, where avatars matched users' expressions live and a 3D parallax effect gave the impression of depth inside a flat screen.
For the lab itself I set up CI/CD pipelines and a private package-manager server, mentored graduate and undergraduate students, and ran multi-day workshops training researchers to build VR studies from scratch.
GIS Developer and Drone Pilot
Orangutans Kutai Project, Bontang, East Kalimantan, Indonesia · September 2009 to November 2018
Three field trips totalling more than 1.5 years, studying wild orangutan behaviour, spatial cognition, and ecology in the rainforests of Borneo.
The computational half is what carried forward. I applied convolutional neural networks to drone and satellite imagery to classify critical orangutan habitat, flew the survey missions myself as pilot and mission planner, and wrote Python packages for spatial data analysis, including a model of GPS error. The work produced peer-reviewed papers and a chapter for Cambridge University Press. Those are listed under publications.