Senior Algorithm Developer — Robotics Team

Kela is a secure, open platform enabling rapid and seamless integration of commercial and military systems in service of Western defense. Designed and battle-tested by Israeli innovators, Kela ensures the West meets today's frontline challenges and evolves with tomorrow's innovations. Backed by leading investors, Kela combines Israeli innovation with global mission impact.
About the Role
Our Robotics department works across unmanned ground vehicles, humanoids, quadrupeds, and drones. We combine classical, modular policy stacks with modern physical AI — integrating and fine-tuning open-source foundation models, and designing and training our own perception, planning, and control modules where nothing off-the-shelf is good enough.
We're looking for a world-class Senior Algorithm Developer to join the team. We need people who think big, move fast, and have a track record that proves they can take an algorithmic problem from a whiteboard sketch to a system running reliably in the field.

Requirements:

Responsibilities

Research and develop algorithms spanning classical methods and deep learning.
Design and build perception, planning, localization, navigation and control modules for real, unstructured environments.
Integrate, adapt, and fine-tune open-source physical AI foundation models.
Own algorithmic pipelines end to end: data, simulation, training, evaluation, deployment.
Apply reinforcement learning and imitation learning to build robust, capable policies.
Build and run simulation environments to train and validate policies before field deployment.
Work closely with systems and software engineers to close the loop between algorithms and real-world performance.
Take ownership of problems end to end, from open question to fielded solution.

Must-have

A proven track record of excellence — top academic background, competitive achievements, elite technological unit service, or high-impact work in industry — and the potential to become one of the strongest in the world in physical AI.
Several years of hands-on experience as an algorithm developer, with real seniority: someone who can own hard problems independently.
Strong foundations in classical algorithms and control, alongside real depth in modern deep learning.
A builder's mindset: comfortable with ambiguity, driven by hard problems, focused on real-world performance.
B.Sc./M.Sc./Ph.D. in Computer Science, Electrical Engineering, Robotics, Applied Math, Physics, or a related field from a leading institution.

Nice-to-have

Hands-on experience with reinforcement learning and/or imitation learning.
Practical experience building data pipelines and simulation environments for training and evaluation.
Strong software engineering skills.
Experience integrating, fine-tuning, or deploying open-source foundation models.
Experience with robotics tooling (ROS/ROS2), simulators (Isaac Sim, Gazebo, MuJoCo), and deep learning frameworks (PyTorch, etc.).

Why Join

Work across UGVs, humanoids, quadrupeds, and drones as the group grows.
Sit at the frontier of physical AI — shaping which approach wins for each problem, classical, learned, or hybrid.
Join a small, high-caliber team where your algorithms reach real robots in the field, fast.
Be part of a fast-growing, well-backed Israeli defense tech company with global impact.