Two new projects, Rongotai Model Train Club (RMTC) and Dailies Notes Assistant, emerge from the Foundation’s new Machine Learning Working Group
August 10, 2025 – The Academy Software Foundation (ASWF), the motion picture industry’s premier organization for advancing open source software development, today announced two new projects: Dailies Notes Assistant and Rongotai Model Train Club (RMTC).
Both projects emerged from the ASWF’s new Machine Learning Working Group (WG), which brings together the machine learning (ML) experts from across the Foundation’s projects and member companies to share expertise and better understand what ML-based tools are needed in the motion picture industry.
“Most studios are considering how machine learning can be used in a film production pipeline but there’s so much hype and uncertainty around it that many are wary and unsure of how to adopt or use the technology, or even how to talk about it,” said Larry Gritz, Chair of the ASWF Technical Advisory Council. “The goal of our Machine Learning Working Group is to create a space where the ML experts in our community can collaborate, incubate and spin up open source projects that use ML technology to empower artists and address the specific needs of our film pipelines.”
The first project to spin out of the Machine Learning WG is a new tool called the Dailies Notes Assistant, which will streamline the dailies process. Dailies Notes Assistant will transcribe dailies meetings, analyze the content using a large language model (LLM), and seamlessly integrate notes directly into Shotgrid for production tracking. ASWF members participating in the development of Dailies Notes Assistant include Industrial Light & Magic (ILM), Sony Pictures Imageworks, DreamWorks Animation, Framestore, and Autodesk.
The second project from the Machine Learning WG, the Rongotai Model Train Club (RMTC), began as an internal project at Wētā FX. Rights and licensing issues, training reproducibility, tooling, datasets, compute and accessibility of model production to artists and Technical Directors are currently significant problems across the industry. To make it easier for artists and Technical Directors to create models and employ ML for niche production problems, RMTC will provide a VFX specific framework for simplifying the production and deployment of ML models and datasets by clearly tracking their provenance and ensuring connection to rights holders. Technical details about RMTC are available here.
“At Wētā FX, we believe the quality we produce is a combination of artistry and enabling technology. With the explosion of ML/AI tools, we continue to focus on empowering artists to tell better stories, but we also want a toolset which provides rights management and tracking for AI similar to how we have managed assets for years. Thus, RMTC was born.” said Kimball Thurston, CTO at Wētā FX. “We hope it helps the community with a common mechanism to integrate ML/AI systems with VFX data formats and concepts, but at the same time, track usage and ensure data and the subsequent models are not used in the wrong context. Looking forward, we hope RMTC provides a vehicle to assist content creators and owners in managing and reporting how AI is involved in the creative process, as well as a bridge between separate dataset provider and content delivery efforts. But we need the whole community to contribute to this conversation, so please join and help.”
To get involved with the Machine Learning Working Group, join the Academy Software Foundation’s #wg-ml Slack channel.
