Job Details

Yandex is looking for a Senior ML Engineer/Researcher for their delivery robot team. Responsibilities include developing World Models, Distributed Training, Model-Based RL & Planning, Representation Shaping, and Safety & Inference. Requires expert knowledge of JAX/PyTorch and...

Yandex delivery robots deliver thousands of orders daily, navigating complex, unstructured urban environments. We are growing rapidly and plan to scale our fleet to 20,000 robots by 2028. We are seeking a Senior ML Engineer/Researcher for our core team to focus on building a fast, interactive world model and large-scale MBRL agent training. What tasks await you: • World Models Development and Scaling You will design and train massive 3D/video tokenizers and backbones based on Diffusion Transformers (DiT), Flow Matching, etc. • Distributed Training You will build pipelines for distributed training of heavy foundation models on our computing cluster. • Model-Based RL (MBRL) & Planning Your task will be to train pure RL and IL + RL policies within a frozen latent simulation of the World Model. • Representation Shaping You will integrate auxiliary losses for 3D detection, segmentation, and tracking perception tasks. • Safety & Inference You will build a reliable safety layer on top of the model's outputs and prepare the entire system for real-time inference directly on the robot's edge devices.

We expect you to: • Have expert-level proficiency in JAX and PyTorch • Possess skills in large-scale distributed training • Have a deep mathematical and ML background • Be capable of writing, generating, and verifying fast, optimized code

It will be a plus if you: • Have worked with Vision Foundation Models, generative video and image models, or lidar cloud synthesis • Have experience in Reinforcement Learning • Have experience with advanced quantization of heavy transformers or diffusion models for edge devices • Have optimized on-board robot inference • Have a background in Autonomous Driving, Motion Planning, or Robotics Why it's good to work with us: We make it easy to grow. If you need to improve your language skills for work tasks, we will organize training and cover 50% of the cost. These are not all the benefits – the full list is here.