Company Description

VYXA is a stealth-stage technology company building a new generation of intelligent products that connect digital intelligence with real-world experiences. We integrate AI, sensing, software and hardware to create intuitive, useful products for everyday life.

We're headquartered in Sydney, with a strategic manufacturing hub in Shenzhen, combining global product vision with rapid hardware innovation. Our team averages over 10 years of experience at leading organisations including DJI, Zero Zero Robotics, Anker, Google, Meta, Canva, Bain Capital, CSIRO and InBody.

Because we're in stealth, early team members have a rare opportunity to shape foundational product and technology decisions.

Role Description

As our core computer vision hire, you'll own the perception stack from research to deployment on real devices in real-world environments. You'll work closely with hardware, software and robotics teams, and your work will directly shape our first products.

What you'll do

- Design, implement and optimise algorithms for real-time detection and tracking of people and small, quick-moving objects in unconstrained indoor and outdoor environments.
- Develop multi-camera calibration, 3D reconstruction and trajectory estimation, and event detection.
- Build human pose estimation and motion analysis capabilities.
- Optimise and deploy models on embedded and mobile platforms under tight latency and power budgets.
- Integrate perception into robotic platforms, including active camera control and person following.
- Build data collection, annotation, evaluation and model validation pipelines.
- Turn research concepts into robust, production-ready systems through rapid prototyping and rigorous experimentation.




- Contribute to system architecture, stay current with advances in computer vision and AI, and help guide technical strategy.

Qualifications

- PhD in Computer Vision, Robotics, Computer Science, Electrical Engineering or a closely related field, or equivalent research-intensive experience.
- Strong foundation in computer vision and pattern recognition, with experience designing and deploying algorithms in real-world environments.
- Deep expertise in at least one of the following: object detection, multi-object tracking, 3D vision or multi-view geometry, or human pose estimation.
- Proficiency in Python and C++, and in PyTorch or TensorFlow, with hands-on experience in modern detection architectures (e.g. YOLO, DETR families).
- Experience deploying models to edge or mobile devices, such as ONNX, TensorRT, Core ML or TFLite.
- Strong fundamentals in algorithm design, data structures, statistical modelling and performance optimisation.
- Ability to work on-site in the Greater Sydney Area, collaborate in cross-functional teams, and communicate complex technical ideas clearly.

Nice to have

- Robotics experience, including perception for autonomous systems, sensor fusion, visual servoing or SLAM.
- Experience with real-time systems, embedded platforms and hardware-software co-design.
- A track record of publications (CVPR, ICCV, ECCV, ICRA, etc.), patents or shipped products in computer vision, robotics or intelligent devices.
- Mandarin proficiency, since we work closely with our Shenzhen hub.

What we offer

- A founding-level role with meaningful equity.
- Competitive salary.
- Real ownership of core technology that ships on real devices.
- The chance to shape a company from its earliest days.

📌 Perception Engineer - Computer Vision (Australia)
🏢 VYXA
📍 Australia