Founding Research Engineer (ML, Perception and Manipulation)

Role overview

This founding engineering position sits at the foundation of robotics AI, building the simulation infrastructure where machines learn perception and manipulation before ever touching the physical world. The team operates as a small, tightly collaborative group taking on ambitious technical problems and valuing ownership, thoughtful risk-taking, and focused execution.

Responsibilities

- Design and extend a simulation platform built specifically for training robotic systems rather than entertainment - Generate large volumes of high-fidelity synthetic data that capture diverse interactions, objects, and environments - Train and evaluate perception and manipulation models that transfer reliably from simulation to physical robots - Shape the developer-facing workflow where models are produced, trained, and pushed onto real hardware without constant rewrites - Take ownership of core platform systems with direct influence on the product's direction - Contribute to a culture that prizes deep learning, careful engineering, and shipping tools that meaningfully reduce wasted effort

Requirements

- Practical background in machine learning, computer vision, robotics, or adjacent disciplines - Familiarity with simulation environments, synthetic data generation, or sim-to-real transfer - Comfort working early in a product's life with significant autonomy and impact - Strong programming and systems engineering skills to build robust, developer-facing tooling - Demonstrated ability to push through hard, ambiguous technical problems

Work setup

The source describes a small founding team environment that prizes ownership, thoughtful risk-taking, and deep focus on impact. Compensation, benefits, and remote policy are not specified in the available material.