Staff Perception Engineer – Robotics
Knightscope, Inc.2 months ago
Sunnyvale, CA, USAStaff+
Base Salary
$240k - $275k/yr
Responsibilities
- Own the full perception stack from multimodal sensor data through 3D detection, scene understanding, persistent tracking, and production deployment.
- Design, train, optimize, and deploy 3D perception models for detection, segmentation, and multi-object tracking using LiDAR, camera, thermal, and radar data.
- Own ML data curation, annotation strategy, model training, TensorRT/CUDA optimization, and real-time onboard edge deployment.
- Build persistent cross-sensor identity tracking for people, vehicles, and license plates across camera handoffs, patrol legs, and time.
- Implement replay-safe, analyst-ready evidence capture including pre/post-event clips, full-frame and cropped media, and structured detection output.
- Explore and integrate vision-language models for scene narration, anomaly reasoning, and long-tail incident understanding.
- Drive architecture, data strategy, and roadmap decisions; mentor engineers across robotics, ML, and systems engineering.
- Design offline and closed-loop evaluations and define safety-critical perception metrics across real-world deployment environments.
Requirements
- 7+ years of hands-on experience shipping perception software to production at an OEM, autonomous vehicle company, physical security platform, or robotics technology company.
- Deep expertise in 3D computer vision, including LiDAR/camera 3D object detection, point-cloud processing, multi-view geometry, and sensor fusion.
- Strong ML fundamentals with experience in CNNs, Transformers, and multi-task architectures.
- Production deployment experience with TensorRT, ONNX, and CUDA optimization.
- Fluency in C++ for real-time systems and Python.
- Experience with ROS2 and simulation environments such as CARLA and Isaac.
- Ability to take models from research into deployed real-world systems operating under operational constraints.
- Preferred: experience with vision-language models, including fine-tuning, prompting, or pipeline integration.
- Preferred: familiarity with Vision-Language-Action or end-to-end policy models for robot behaviors.
- Preferred: experience with ALPR, face recognition, or ReID systems in deployed security or automotive contexts.
- Preferred: familiarity with evidence integrity, chain-of-custody media, or tamper-evident capture pipelines.
- Preferred: publications or open-source contributions in 3D computer vision, embodied AI, or autonomous systems.
- Preferred: MS or PhD in Computer Science, Robotics, Electrical Engineering, or a related field.
Benefits
- Medical, dental, vision, 401(k), paid time off, and stock options.
- Full-time, on-site work at the Sunnyvale headquarters; no relocation provided.
Categories
ML EngineeringRobotics