Senior AI Engineer

The role

We’re looking for a Senior AI Engineer to help build the intelligence layer behind ZYGOS.

This role is focused on agentic AI, reinforcement learning, and Physical AI. You will design systems that allow robots to reason, plan, learn from simulation and real-world feedback, and execute tasks in dynamic environments.

You will work on AI agents that connect perception, language, planning, robot skills, and control. The goal is not to build demos that look good once. The goal is to build reliable Physical AI systems that can be deployed on real robots and improve over time.

This is a senior role with high ownership. You will help define ZYGO’s AI architecture, build core components yourself, guide technical decisions, and work closely with robotics, simulation, cloud, and product teams.

What you will do

• You will design and build AI systems that can understand goals, select the right robot skills, plan multi-step actions, and recover when the environment changes.
• You will develop reinforcement learning pipelines for robotics tasks, using simulation, real-world data, and evaluation loops to improve robot behaviour.
• You will work on Physical AI models that combine perception, language, planning, memory, and action into deployable robotic systems.
• You will build training, fine-tuning, evaluation, and deployment pipelines for models used in robot reasoning, navigation, manipulation, inspection, and task execution.
• You will collaborate with robotics engineers to connect AI models with ROS2, robot skills, simulation environments, and real hardware.
• You will help define how ZYGOS uses LLMs, VLMs, world models, reinforcement learning, imitation learning, and classical robotics methods together in one production system.
• You will turn research concepts into working software that can survive real-world constraints such as latency, uncertainty, limited compute, noisy sensors, and safety requirements.
• You will create benchmarks, datasets, simulation scenarios, and automated tests to measure whether AI behaviour is improving in a repeatable way.
• You will mentor other engineers and help set the technical direction for AI inside ZYGO.

What you bring

• You have strong practical experience in AI engineering, machine learning, or robotics AI, with a clear ability to move from research ideas to production systems.
• You understand agentic AI architectures and know how to build systems that combine reasoning, planning, tool use, memory, and execution.
• You have experience with reinforcement learning, imitation learning, robotics simulation, or policy learning for embodied agents.
• You are comfortable with modern AI tooling such as PyTorch, transformers, multimodal models, vector databases, evaluation frameworks, and model deployment pipelines.
• You understand robotics constraints and can work with real-world systems where perception is noisy, actions have consequences, and reliability matters.
• You write clean, reliable Python code and are able to build production-grade AI pipelines, not just notebooks.
• You can work with simulation environments and understand the importance of sim-to-real validation.
• You care about measurement. You know that an AI system is only useful when its behaviour can be evaluated, compared, improved, and deployed safely.
• You are senior enough to make architectural decisions, challenge assumptions, and take ownership of complex technical problems.
• Experience with ROS2, Isaac Sim, MuJoCo, Gymnasium, RLlib, Stable-Baselines, diffusion policies, VLA models, Jetson deployment, ONNX, TensorRT, or CUDA is a strong plus.