Position Overview
As a Principal Research Scientist - Datasets at Autodesk Research, you will be doing fundamental and applied research that will help our customers imagine, design, and make a better world.
We are a team of scientists, researchers, engineers, and designers working together on projects that range from learning-based design systems, computer vision, graphics, robotics, human-computer interaction, sustainability, simulation, manufacturing, architectural design and construction.
This role will report to the Research Dataset Creation & Annotation Manager in the AI Lab.
Responsibilities
1. Development of specialized datasets to evaluate, fine-tune, and train large 2D & 3D models for Design & Make
2. Design engineering benchmarks and evaluation frameworks, including task formulation, metric selection, and validation protocols, to assess model and dataset quality
3. Drive data collection projects from start to finish by gathering requirements, defining success metrics, and adjusting to the dynamic requirements of AI Research
4. Review relevant AI/ML literature to identify emerging methods, technologies, and best practices
5. Explore new data sources and discover techniques for best leveraging data
Minimum Qualifications
6. A Master's or PhD in a field related to AI/ML such as: Computer Science, Mathematics, Statistics, Physics, Linguistics, Mechanical Engineering, Architecture or related disciplines
7. Good background in statistical methods for Machine Learning (e.g. Bayesian methods, HMMs, graphical models, dimension reduction, clustering, classification, regression techniques, etc.)
8. Familiarity with Deep Learning techniques (e.g. Network architectures, regularization techniques, learning techniques, loss-functions, optimization strategies, etc.)
9. Strong coding abilities in Python and/or C++
Preferred Qualifications
10. Experience in Product Design & Manufacturing or other Autodesk domains
11. 2D & 3D Generative AI
12. LLMs and Natural Language Processing
13. Multi-modal deep learning and/or information retrieval
14. Computational geometry and geometric methods (e.g. shape analysis, topology, differential geometry, discrete geometry, functional mapping, geometric deep learning, graph neural networks)
15. Publication track record at relevant conferences
16. Significant post-graduate research experience, or 5 or greater years of work experience