Head of Synthetic Data Generation

We’re building an AI assistant for hardware designers. Our mission is to help millions of engineers iterate through designs 1000x faster.

We’ve developed geometry- and physics-driven foundation models, and now we’re looking for builders who thrive at the frontier of product, data, and intelligence.

Are you the kind of person who likes to turn complexity into clarity? Do you think in pixels, meshes, physics engines, and procedural chaos? We're looking for a creative systems-minded Head of Synthetic Data Generation to lead our charge into the synthetic unknown.

This isn’t about building imaginary game worlds. It’s about capturing and remixing the real, grounded, and highly constrained world of hardware components, mechanical assemblies, and design intent — then generating synthetic data that helps our models deeply understand it all.

You’ll partner closely with ML engineers, designers, and simulation experts to fill gaps, surface edge cases, and expand model generalization through intelligent data creation.

This is a playground for inventive minds who love clean abstractions, weird edge cases, and building things that actually work.

• Invent and Build: Design end-to-end pipelines to generate synthetic 2D, 3D, or parametric data that reflects real-world hardware use cases — and pushes model capabilities forward.

• Stay Grounded, Think Bold: Work within real-world constraints (tolerances, physics, manufacturability) while embracing creative approaches to finding and leveraging data both from the physical and digital worlds.

• Be the Voice of Synthetic: Define the synthetic data roadmap and evangelize the power of simulation, generative tools, and data-first development across the org.

• Push the Edge: Mix procedural generation with NeRFs, diffusion models, generative rendering, and other bleeding-edge methods. This is your sandbox .

• Automate and Scale: Build the infrastructure to generate high-quality, high-diversity synthetic datasets that integrate cleanly with training pipelines.

• Measure what Matters: Create evaluation frameworks that link synthetic data interventions to real model performance improvements.

• Lead the Team: Build and mentor a team of engineers, creators, and technical experts who want to make machines smarter by making fake stuff really well.
  

• Deep experience in simulation, procedural generation, or synthetic data at scale (think: AVs, robotics, vision systems, 3D modeling, etc.).

• Strong familiarity with procedural generation, parametric modeling, or simulation (e.g., Blender, CAD kernels, custom mesh tools).

• Fluency with modern ML workflows: PyTorch, dataset tooling, training loops, evaluation frameworks.

• Strong creative problem-solving skills —you know when to write code, when to mutate geometry, and when to invent synthetic failure cases to close a learning gap.

• Leadership experience: you’ve built and led multidisciplinary teams or at least acted as the glue between engineers, artists, and ML researchers.

• You love a greenfield challenge and see ambiguity as a canvas, not a blocker.
  

• Background in game dev, computer graphics, reinforcement learning, or 3D CAD.

• Experience with STEP, STL, OBJ, or other hardware design formats.

• Prior experience shipping synthetic data into production ML loops.

• An eye for aesthetics and weirdness. (Seriously—sometimes the model needs weird data.)
  

• You’ll help build the future of AI-powered hardware design.

• You’ll work on hard problems with real-world consequences.

• You’ll operate at the intersection of geometry, machine learning, and design — and help define a new category of tools.

• You’ll join a team that values clarity, experimentation, and creativity with constraints.

• Competitive salary, equity, and benefits — plus the chance to define a new category of tooling from the ground up.