Computer Vision Engineer

Mkind is an early-stage robotics startup on a mission to automate the hardest and most painful manual industrial tasks. We're building a fully flexible, intelligent robotic work cell for surface finishing, welding, and coating — the jobs that matter most in modern industry. Small, ambitious team based in Munich, and we move fast.

A foundational role: you'll be the person who gives our robot its eyes. Without high-quality perception, the robot can't understand the surface it's working on, can't adapt to variation, and can't operate autonomously in unstructured industrial environments.

You'll work directly on the core tech — developing real-time vision pipelines, point cloud processing, and surface analysis algorithms that directly drive robot behavior. You'll be defining the architecture, not just implementing it.

• Design and implement computer vision pipelines for real-time surface analysis and defect detection.
• Develop 3D perception systems using depth cameras, laser sensors, or LiDAR for point cloud processing.
• Build and maintain calibration routines for the perception sensor stack.
• Train and deploy deep learning models for object detection, pose estimation, segmentation, and defect classification.
• Benchmark and optimize inference performance for embedded / edge deployment.
• Collaborate with the robotics team to translate visual outputs into actionable robot control signals.
• Define and iterate on the sensor stack (camera selection, mounting, illumination).

• Strong programming skills in Python and/or C++.
• Hands-on experience with OpenCV, PCL, or similar CV / 3D processing libraries.
• Solid understanding of camera models, calibration, and stereo / depth imaging.
• Experience with deep learning frameworks (PyTorch or TensorFlow) for vision tasks.
• Ability to work in a fast-moving startup environment with high ownership.
• Degree in Computer Science, Electrical Engineering, Robotics, or a related field.

• Experience with robotic systems (ROS / ROS 2).
• Background in industrial inspection, quality control, or surface analysis.
• Familiarity with CUDA or other GPU acceleration for real-time inference.
• Prior experience in an early-stage startup or research environment.

Location: Munich, Germany (on-site, 5 days/week)
Equity participation as an early team member.