Clera Founding Engineer, Computer Vision & Robotics at Clera building perception systems for manufacturing quality inspection. Responsible for developing and deploying real-time computer vision models.
Responsibilities
Join an early-stage physical AI company building computer vision systems for manufacturing quality inspection. As the first engineer, you will take perception from experimentation to reliable factory deployments, working closely with the founding team.
Build perception systems for detection, classification, tracking, temporal understanding, and 3D vision.
Develop first-person perception that handles camera motion, changing viewpoints, occlusion, motion blur, and variable lighting.
Improve models, data, and evaluation by turning failures in real deployments into system improvements.
Build geometric vision capabilities, including calibration, localization, pose estimation, and coordinate transformations.
Train and debug PyTorch models using synthetic and real-world data.
Optimize low-latency inference and deploy systems on NVIDIA Jetson using TensorRT.
Work on factory floors and build reusable perception infrastructure from deployment learnings.
Qualification
Strong PythonExperience building detectionBackground in computer visionExperience with SLAM or VIOExperience with depth estimationExperience with CUDA or TensorRTExperience with ROS2
Required
At least 3 years of experience building and deploying computer vision, robotics perception, or applied ML systems beyond a notebook or lab.
Strong Python, PyTorch, and OpenCV skills, with experience shipping real-time perception systems in physical-world conditions.
Experience building detection, classification, or tracking models and working with geometric vision.
Experience deploying perception in production, optimizing for edge or low-latency inference, and working directly with end users or in the field.
Background in computer vision, robotics, industrial robotics, AR/VR, ADAS perception, or physical AI.
Experience with synthetic data and Sim2Real, SLAM or VIO, depth estimation, 3D reconstruction, CUDA or TensorRT, ROS2, robot manipulation, or robot learning is beneficial.
Ability to own an underspecified physical-world problem end to end.
Preferred
Experience with synthetic data and Sim2Real
Experience with SLAM or VIO
Experience with depth estimation, 3D reconstruction
Experience with CUDA or TensorRT
Experience with ROS2, robot manipulation, or robot learning
