The company builds hardware foundation models for mechanical engineering. Its GPU-accelerated design models let engineers generate and compare hundreds of complete mechanical assemblies in minutes, speeding up the invention of new hardware.
The models learn from a company's past CAD designs and produce editable assemblies from engineering requirements. They include standard and custom parts and respect manufacturing constraints, helping customers ship their next hardware generation sooner.
The founders include a Carnegie Mellon professor of ECE and robotics, a Forbes Under 30 serial hardware entrepreneur, and a Harvard PhD who was a scientist at Amazon Robotics. Together they have written over 65 papers on hardware design algorithms. Team members hold PhDs from Imperial, TU Wien, and UCL, and work in computational geometry, computer vision, ML, optimization, simulation, and CAD.
Responsibilities
Build the interface engineers use to describe, generate, inspect, and refine designs.
Turn natural-language and multimodal inputs into structured engineering specifications.
Work on frontend development, agent and tool orchestration, and frontier model adaptation and evaluation.
Connect the product with CAD, geometry, and backend systems.
Evaluate, adapt, and improve the underlying models so the product behaves reliably.
Create product experiences that are reliable and intuitive around the models.
Requirements
Strong product engineering skills and a record of building polished, user-facing applications.
Understanding of how LLMs behave beyond prompting and basic API calls.
Ability to turn vague user intent into structured, executable workflows.
Experience designing agentic systems that use tools and backend services reliably.
Sound judgment on model selection, evaluation, and failure analysis, including when adaptation or fine-tuning is needed.
Clear product and technical judgment when designing new kinds of human-AI interaction.