Lead Perception Engineer

Fruition Group US; California (MO); On-site; USD 120,000 - 160,000; Full time

Job summary

Fruition Group US is building the perception stack for a new wave of robotics products focused on manipulation and robotic arms. We aim to create perception that sees, understands, and acts in unstructured real-world environments, with early team development and real-world product readiness in perception systems for manipulation.

The role requires strong computer vision and deep learning experience, production deployment to edge or embedded platforms, and a background in 3D perception, SLAM, or

Qualifications

• Experience in computer vision and deep learning applied to robotics.
• Production deployment of models to edge or embedded platforms.
• Background in 3D perception, SLAM, or sensor fusion.
• Strong Python and C++ programming skills.
• Experience with large-scale real-world datasets.

Responsibilities

• Design, train, and deploy perception models for object detection, tracking, segmentation, and 3D scene understanding.
• Set technical direction for perception and control of robotic arms in real-world environments, with a track record on shipped systems.
• Push perception from prototype to production on real hardware, addressing latency, lighting, occlusion, and edge case challenges.

Skills

Computer vision

Deep learning

PyTorch

Python

C++

3D perception

SLAM

Sensor fusion

Edge deployment

Job description

We're building the perception stack for a new wave of Robotics product, purely focused across major innovation within robotics manipulation and robotics arms. We are building the kind that see, understand, and act in unstructured, real-world environments. Early team and real world products : Perception systems for manipulation, including sensing, understanding, and reasoning.

Responsibilities:

• Design, train, and deploy perception models for object detection, tracking, segmentation, and 3D scene understanding.
• Set technical direction for how the arm perceives and acts on the physical world, backed by a track record of making that call correctly on at least one shipped system, in close partnership with controls and applied research.
• Push perception from prototype to production on real hardware, applying hard-won lessons about what breaks when a model leaves the lab — latency, lighting, occlusion, edge cases no dataset captures.

What We Need:

• Strong experience in computer vision and deep learning (PyTorch or equivalent).
• Production experience deploying models to edge or embedded platforms.
• Background in 3D perception, SLAM, or sensor fusion.
• Proficiency in Python and C++.
• Experience working with large-scale real-world datasets.