Backflip is looking to solve the problem of making 3D CAD as intuitive and fast as sketching on paper by building AI-first tools, specifically by training and deploying frontier multimodal models for 3D CAD.
Requirements
- 2–7 years of experience in Data/ML infrastructure with strong CS fundamentals
- Direct, hands-on experience training, and deploying models in production
- Proficiency in Python and PyTorch
- Experience taking ML projects from prototype to production
- Deep understanding of the ML lifecycle and infrastructure
- Experience with CAD, 3D geometry, graphics, differentiable programming, or large-scale data pipelines
Responsibilities
- train and deploy frontier multimodal models for 3D CAD
- develop novel representations for complex 3D data
- train models that generate structured design outputs
- build scalable data pipelines
- deliver low-latency inference in production
- rapidly prototype and productionize features
- collaborate closely with AI research, product, and design to define what state-of-the-art generative design looks like
Other
- 2–7 years of experience
- In-person in SF
- Collaboration with AI research, product, and design teams
