AI Research Scientist – Structured & Spatial Modeling
| Company | Autodesk |
|---|---|
| Location | Montreal, QC, Canada, Winnipeg, MB, Canada, Toronto, ON, Canada, San Francisco, CA, USA, Calgary, AB, Canada, Charlottetown, PE, Canada, Regina, SK, Canada, Vancouver, BC, Canada, Halifax Regional Municipality, NS, Canada, Fredericton, NB, Canada |
| Salary | $147300 – $254100 |
| Type | Full-Time |
| Degrees | PhD |
| Experience Level | Senior, Expert or higher |
Requirements
- A PhD in a field related to AI/ML such as: Computer Science, Mathematics, Statistics, Physics, Linguistics, Mechanical Engineering, Architecture or related disciplines
- Publications in venues such as NeurIPS, ICLR, GECCO, MLSys, JMLR, ICML, IROS, EMNLP, MICCAI, Medical Image Analysis, IEEE TMI, ICDAR, CVPR or other high impact ML conferences and journals
- Strong background applying Deep Learning techniques, particularly generative models and spatial reasoning (including implementing custom architectures, optimizing model performance, developing novel loss functions, and deploying production-ready solutions)
- Strong coding abilities in PyTorch
Responsibilities
- Develop and lead research on novel ML models for structured generation and multi-modal understanding of complex visual and spatial content
- Review relevant AI/ML literature to identify emerging methods in generative models and autoregressive architectures
- Explore new data sources and discover techniques for leveraging structured representations in design workflows
Preferred Qualifications
- Experience in the Architecture, Engineering, and/or Construction domains, including expertise with industry-specific data formats (e.g., BIM models, IFC files, AEC Contract Documents and Drawings) and structured data representation in AEC workflows
- Experience in specialized domains requiring sequential or structured generation of complex spatial annotations, such as medical imaging (autoregressive segmentation, sequential annotation), document layout analysis (technical diagrams, forms, scientific papers), or cartography/GIS (map annotation, symbol placement)
- Experience with LLMs and autoregressive models for structured representations such as code generation, molecular/biological discovery, or other domain-specific sequence modeling
- Multi-modal deep learning and/or information retrieval experience
- Significant post-graduate research experience, or 5 or greater years of work experience (actual job title/position will be commensurate to experience)