Pipeline development with a software lean, while Mindshow prepares to offer their in-house product to outside businesses / clients.

• Created a microservice to convert FBX files to Mindshow assets using AWS Lambdas, SQS, EC2s, S3, and Python

• Reduced time of building Maya shots from mocap data by only baking down animation keys for shot frame ranges, decreasing build time by best case 75% (20 min to 5 min)

• Worked with the UI/UX and Unity teams to build the front-end of a registration tool using PyQt, CSS, Figma, and REST API that takes field inputs, grabs the machine's MAC address, and registers the machine to the database

Pipeline development for a VR-based startup, with a focus on hybrid (on-prem + cloud) render farm setup using ASWF's OpenCue.

• Designed and implemented a wrapper around OpenCue, an open-source render management tool, using Python, Docker, subprocesses, and OpenCue API to make farm job submissions easier to launch from our existing pipeline

• Worked with DevOps and IT to set up a hybrid on-premise and on-cloud render farm using OpenCue, and AWS EC2s to dynamically scale up render nodes, saving costs and allowing bursting to cloud during peak render times

• Converted manual processes in Maya, such as renders, playblasts, and sequence builds, into farm jobs using Python and OpenCue to decrease artist and production down time, saving them around 3 hours a week

• Automated Aspera package ingestion using Aspera Faspex, Watchdog, and ShotGrid Event Daemon plugins to bypass manually downloading vendor submissions from Aspera Connect, saving Production around 1.5 hours a week

Web application and architecture development for client-side production of movies and television for studios under Disney banner, such as Marvel Studios, Disney Motion Picture Productions, and Searchlight Pictures

• Updated file transfer monitoring web application using Python, Django, HTML, CSS, and Javascript to onboard Finishing team, expanding app's user base by 20% and improving visibility

• Created in-house Jira Python API and incorporated a microservice framework using Celery, RabbitMQ, dead letter exchanges for better error handling when Jira server goes down, decreasing bug reports and debugging time by 50%

• Built Docker container and image to spin up Deadline servers using Docker, Git CI/CD, and pipenv in order to quickly spin up and shut down local test instances, speeding up testing feedback loop by 100%

• Collaborated with group of 5 engineers to prototype auto-deployment workflow for Deadline farm workers using Ansible Playbook, Docker, Git CI/CD, and Linux VMs so 40 render nodes could be batch updated and configured

Pipeline setup, tool development, and support for ILM's Sydney studio

• Updated custom import and export nodes in Houdini using Houdini nodes, Python, and VEX to properly read and write custom Crowd objects from and to Alembic cache files, allowing Generalists to use Houdini for Crowd work on Dungeons & Dragons: Honor Among Thieves

• Crewed as sole Pipeline TD in StageCraft Brain Bar for on-set filming of Thor: Love & Thunder, creating tools, troubleshooting, and serving as backup operator to ensure technoloy runs smoothly during shoots

Credits:

• The Mandalorian, Season 2 (2020)

• Thor: Love and Thunder (2022)

• Dungeons & Dragons: Honor Among Thieves (2023)