Dream3D
AI-powered 3D Design
ML Research Engineer - AI-Powered 3D Design Tool in Brooklyn, NY
About the role
Dream3D is building a new generation of 3D tools for design, animation and gaming. We’re tackling hard problems across computer graphics, AI and UI/UX design to build the future of 3D creation.
We’re a team of designs/creative technologists, and machine learning experts passionate about tool making and exploring our relationships with intelligent systems. We’re backed by great investors like SNR, Twenty Two VC, Y Combinator, and are looking to grow our founding team.
If this resonates, we want to meet you.
About Us
We’re both second-time founders -
- Tony Francis: Started Inokyo, building AI-powered retail stores (like Amazon Go). Led machine learning platform at Clubhouse.
- Justin Woodbridge: Started Bayes, a design-focused data visualization tool. Acquired by Airtable in 2021.
About the Role
We’re looking for an exceptional machine learning engineer to join our founding team. You’ll be working directly with Tony to implement state-of-the-art research papers, do your own research and when the moment is right, train foundational models.
Responsibilities:
- Develop state-of-the-art generative models for 3D content creation
- Read, implement and improve the latest papers in this field
- Rapid prototyping of ideas, approaches, methods etc.
Ideally you:
- Have 3+ years of machine learning engineering/research experience
- Track record of implementing research papers in PyTorch
- Solid understanding of state-of-the-art generative models (VAEs, GANs, Autoregressive, Diffusion, etc.)
- Solid understanding of 3D math and modern 3D graphics
- Familiar with differential rendering
- Knowledge of computer graphics fundamentals (rendering, geometry processing, etc.)
We’re big believers that small, in-person teams are the ones that create the future. We work out of our Brooklyn, New York office, but are flexible with hybrid work schedules. Ideally you are based in NYC.
About Dream3D
Dream3D is an AI-powered 3D design tool. It lets designers instantly create and render 3D scenes with AI.
Designers can build a complete 3D scene by asking for what they want – “a jet ski in front of a building”, and then render a final image by describing how it looks – “photorealistic yellow jet ski, evening sun, 80s magazine ad”.
We see ourselves as the next generation Autodesk, focusing on professional design workflows instead of building for hobbyists. Unlike current generative tools, we give users the fine-grain control they need for professional projects.
Our goal is to become a complete 3D suite for animation and creating full 3D worlds. We’re building the generative engine for video games, AR/VR and even industrial simulation tasks like training embodied AI.
