Engineering leader with over 20 years building technology for film and 3D production, currently leading multiple teams across engineering, QA and product. I prototype to make ideas concrete, then build the buy-in that makes them stick. My work has shaped practices well beyond my own teams, from a refinement model adopted across Katana to a company-wide guideline for AI-generated code.
Selected Impact
- Delivery: shipped eight feature releases across Nuke and Katana while splitting a team and absorbing a restructure that doubled my scope.
- Engineering effectiveness: 3-4x more tickets refined per session, typical code review under two days in two teams, and local builds cut from ~20 min to ~2 min.
- Organisational influence: practices adopted beyond my teams, including a refinement model used across Katana, AI research ways of working used by four teams, and a company-wide guideline for AI-generated code.
- Leadership development: grew two engineers into the engineering managers who succeeded me, promoted two to senior, and coached two peer managers. At the CTO's request, I mentor a senior engineer in another product toward management.
Skills
- Leadership: Multi-Team Engineering Leadership, Organisation Design, Delivery Systems, Coaching, Hiring
- Effectiveness & AI: Developer Productivity, AI-Native Engineering, Agent-Ready SDKs
- Technical: C++, Python, Build Systems, GPU Computing, Machine Learning, OpenUSD
Experience
Foundry, London
Engineering Manager, Katana (Lead Software Engineer) | Jul 2025 - Present
- Lead two engineering teams (~13 people across engineering, QA and product) on Katana, a large C++/Python 3D application used by major film studios. Shipped Katana 8.5 and 9.0 plus regular maintenance releases, and absorbed a restructure that doubled my scope without missing a release. Promoted an engineer to senior.
- Improved delivery predictability in both teams: brought typical code-review turnaround under two days with a review-tagging system, and made feature reviews happen every sprint with async video demos instead of meetings.
- Raised tickets refined per session 3-4x with a Feature Facilitator role, adopted voluntarily by the other two Katana teams: a rotating engineer who prepares upcoming work with product and architecture one sprint ahead.
- Defined Katana’s approach to customers building plugins with AI coding agents, after identifying that our docs and examples were not usable by agents. Wrote the agent-ready SDK strategy, validated it with major studios, and saw its objectives adopted by a director-led AI initiative.
- Drive AI-native engineering beyond my teams: originated Foundry’s guideline for labelling AI-generated code, and co-lead an organisation-wide AI software factory initiative with two peer leaders, now running experiments.
- Removed long-standing productivity bottlenecks across the product: compiler caching (local builds ~20 min to ~2 min); zero-setup Python IDE and debugger support, closing a 15-year tooling gap; one-click profiling; and an engineering-insights platform over GitLab and Jira.
Engineering Manager, AI Research (Lead Software Engineer) | Jul 2022 - Jul 2025
- Set roadmaps with product, led planning, and presented the group’s work to the executive team at quarterly reviews; designed the research-engineering hiring framework, later adopted by other groups.
- As the group’s first engineering leader, split one team into two (~9 people) as demand grew and established ways of working later adopted by four teams.
- Shipped production machine-learning features across six feature releases of Nuke (14.0 to 16.0), the industry-standard compositing software, including major expansions of its trainable ML tools and model inference.
- Grew two senior engineers into the engineering managers who succeeded me in leading the group, one through a year of coaching, one by recognising her readiness and creating room to step up. Promoted another to senior.
Senior Software Engineer, Real-Time Research | Sep 2020 - Jul 2022
- Most senior engineer on Nuke Stage, a real-time virtual-production product, while its direction was still open: prototyped GPU kernels, application architecture, camera tracking and networking. The shipping product built on this work is now used on live film productions.
- Built the prototype of Nuke’s 3D system on OpenUSD (Pixar’s open 3D scene standard) with the lead architect; it was greenlit and matured into what ships today.
DNEG, London
Software Engineer, R&D | Aug 2019 - Sep 2020
- Designed and built the asset-resolution layer of the studio’s OpenUSD platform, which locates and loads assets across productions.
- Built pipeline tools used on Tenet, No Time to Die, and Fast & Furious 9.
Animal Logic (now Netflix Animation), Sydney
Software Engineer, R&D | Feb 2016 - Aug 2019
- Created an in-house authoring framework that changed how the entire character-rigging team worked, used on Captain Marvel, The LEGO Movie 2, and the Peter Rabbit series.
- Contributed to one of the first adoptions of OpenUSD at feature-film scale outside Pixar, solving source-of-truth sync in a pipeline mid-migration and helping design the approach that followed.
VFX Industry, Multiple Studios
Rigging Supervisor / Lead & Senior Character TD | 2005 - 2016
- Rigging Supervisor at Kandor Graphics, leading the character-rigging team through a feature film; Senior Character TD on a decade of released productions.
Education
Universidad Central, Chile
B.Sc. in Construction Engineering | 2000 - 2005
Passed with Distinction