Applied Scientist - State Estimation & Integration

About the role

The company, whose mission is to solve fundamental challenges in AI and robotics, is looking for Master's and PhD graduates to join its team full-time as the bridge between algorithmic research and on-robot deployment. The role sits at the intersection of state estimation, sensor calibration and high-performance systems engineering, building the computational foundations and state estimation suites for multi-modal sensor systems (RGB cameras, LiDARs, IMUs) so they run seamlessly on physical hardware.

Key responsibilities

• Develop and integrate robust, real-time state-estimation pipelines for multimodal sensor systems.
• Write robust, low-latency modern C++ code for real-time execution on robotic hardware, optimising for embedded platforms such as the Jetson family.
• Apply classical optimisation (e.g. factor graphs) and learning algorithms to real-world robotics problems.
• Profile, benchmark and eliminate performance bottlenecks in resource-constrained Linux environments.
• Design and build clean, scalable software components using modern software engineering practices and ROS 2.
• Work with researchers to provide a solid foundation for advanced robotic systems, making complex algorithms easy to use and deploy.

Required profile

• Master's degree or a completed (or soon-to-be-completed) PhD in Robotics, Computer Science, Computer Vision, Computational Science & Engineering or a related field; early-career and junior candidates with a few years of experience are welcome.
• Proven software engineering experience, such as industry internships, significant open-source contributions or complex project work.
• A strong interest in building tools that support both robotics research and real-world deployment.

Required skills

• Mandatory proficiency in modern C++ and Python, including system architecture, data structures and memory management.
• Solid understanding of state estimation and optimisation techniques, with familiarity in factor graph optimisation (ideally with GTSAM).
• Bonus: multi-sensor state estimation, VIO or SLAM; GPU programming, CUDA or embedded ML inference; open-source robotics projects; computational physics or rendering pipelines such as Gaussian splatting.