Research Engineer

Company OverviewEach year, 2.5 trillion hours are spent on household chores. At Matic, we’re on a mission to recapture that lost time, and we’re doing it by revolutionizing home robotics.Our first product, also called Matic, is a Wall-E-esque floor cleaning robot. We've built what amounts to "full self-driving in the home” with real-time 3D mapping, adaptive path-tracking, and a precise semantic understanding of the home. Our breakthroughs in spatial AI allow Matic to work reliably in real homes, using only RGB cameras and neural networks running on-device.Privacy FirstWhat happens in the home, stays in the home. Our robots are private by design, with all data processing performed by the robot itself, not in the cloud.Our ApproachBefore the iPhone, consumers adopted several distinct devices; cell phones, PDAs, and portable music players each served a particular need. We believe in a similar progression for home robotics, starting with single-purpose robots and building iteratively toward more complex capabilities over time.Our CultureMatic is a tight-knit and collaborative team, singularly focused on building products our customers will cherish. We're ultra-hardworking people committed to solving tough problems that save precious time and energy.About The RoleWe're looking for a motivated perception engineer to join us on the ground floor, with a runway for huge and immediate impact.What You'll DoTranslate open-ended, real-world problems into well-defined machine learning problemsDesign, build, and code computer vision-based algorithmsBuild novel neural network architecturesDesign and implement robust 3D geometry-based vision algorithmsBuild scalable machine learning infrastructure to train on large datasets and simultaneously be reproducibleAnalyze results, continuously update and improve accuracy & speed of various algorithmsDeploy algorithms & machine learning pipelines at scale on the edge devices with limited compute without sacrificing performanceWork to enable Level 5 autonomy in consumer roboticsCross-collaborate with Hardware, Electrical, and Systems teams to bring product vision to lifeWhat We Look ForRobust machine learning background in computer vision and perceptual tasks, including 3D perception, multi-class object detection, 3D maps, and advanced robotics controlsTrack record of product-focused research applied to real-world applicationsFluency with Rust (or C++) and Python (PyTorch or similar tools)Ability to write clean, reliable code that is easily maintainedWe'd love to hear from you if...You are genuinely motivated to help those around youYou are passionate about learning outside of your normal comfort zonesYou love diagnosing complex technical issuesYou are excited to do great work