1 day ago
Toronto, Canada or Montréal, CanadaSenior
Responsibilities
- Review state-of-the-art media- and motion-focused machine learning publications and technologies.
- Evaluate feasibility and generalizability and report findings to project and program leads.
- Implement cutting-edge machine learning results and improve existing models.
- Develop specific and generalized large models for motion prediction across multiple modalities.
- Create end-to-end workflows using data-driven techniques and motion-editing technologies.
- Collaborate on research projects within a globally distributed team.
- Build experimental pipelines and prototypes to support research.
- Process and prepare data at scale.
- Train models and perform hyperparameter tuning.
- Analyze model errors and design mitigation strategies.
- Maintain scalable implementations and high engineering standards for shared code repositories.
Requirements
- Master’s or PhD in AI/ML, computer science, mathematics, statistics, physics, or a related discipline.
- Proficiency in deep learning techniques, including network architectures, regularization, learning techniques, loss functions, and optimization strategies.
- Deep knowledge of statistical machine learning methods, including Bayesian methods, HMMs, graphical models, dimensionality reduction, clustering, classification, and regression.
- Experience with multiple assisted coding environments and scalable assisted code-generation practices.
- Knowledge of computational time-series data analytics and prediction regimes.
- Proficiency with scalable compute, bandwidth, storage infrastructure, and associated orchestration layers.
- Proficiency with Python, TensorFlow, PyTorch, and other standard deep learning frameworks.
- Strong understanding of fundamental computer science algorithms and their scaling behavior.
- Practical experience applying machine learning algorithms and libraries to business problems.
- Preferred: production experience and creative workflow experience in entertainment or adjacent industries.
- Preferred: demonstrated collaboration with creatives and producers.
- Preferred: familiarity with media pipelines and practical data modeling.
- Ability to decompose complex problems, work independently, document code and software architecture, communicate effectively, and adapt to ambiguity.
- Passion for entertainment and experience using creative software are preferred qualities.
Benefits
- Comprehensive benefits package may be included as part of the compensation package.
- The team is globally distributed, with key locations in Toronto, Montreal, San Francisco, and Vancouver.
- The role may require in-person onboarding and/or in-person identity verification.
Categories
About Autodesk
Autodesk builds design and engineering software used in architecture, construction, manufacturing, and media and entertainment. Its subscription-based portfolio includes AutoCAD, Revit, Fusion 360, Maya, and 3ds Max, along with cloud collaboration and simulation services. Founded in 1982 and headquartered in San Francisco, the public company trades on NASDAQ as ADSK and serves customers ranging from AEC firms to product designers and film and game studios.