Senior Machine Learning Engineer

About this role:Pickle is on the hunt for a dynamic and driven Senior Machine Learning Engineer to push the boundaries of robotic AI and machine learning in industrial automation. This role is deeply R&D-focused, working on prototyping, developing, and refining machine learning models that power our next-generation robotic unloading solutions. You will explore, design, and implement cutting-edge AI/ML algorithms for robotic perception, decision-making, and adaptive control in warehouse automation environments.

What you'll Accomplish:

• Research, prototype, and develop AI-driven models for robotic manipulation, grasping, and object interaction.
• Design and optimize Classical ML / deep learning and reinforcement learning (RL) models for robotic control and decision-making.
• Develop simulation-driven ML pipelines, integrating synthetic data, real-world sensor data, and reinforcement learning.
• Work with different learning models, including Large Language Models (LLMs), transformer-based architectures, and generative AI to enhance robotic intelligence.
• Collaborate with robotics engineers, software developers, and mechatronics teams to integrate ML models into real-world robotic systems.
• Design and implement scalable ML frameworks for real-time robotics applications.
• Conduct model evaluation, benchmarking, and performance tuning to improve accuracy, robustness, and generalization.
• Develop and maintain data pipelines, including sensor data collection, annotation, and preprocessing for AI model training.
• Experiment with foundation models and generative AI for robotics applications.
• Work with physics-based simulators (Isaac Sim, Mujoco, PyBullet, Gazebo, DRAKE) to train and validate ML models.

Who you are:

• 5+ years of experience in machine learning, deep learning, or reinforcement learning for robotics.
• Hands-on experience with robotic perception, motion control, and adaptive decision-making models.
• Strong background in ML model development, training pipelines, and real-world deployment.
• Experience in R&D environments, rapidly prototyping and iterating on AI models.
• Familiarity with robotics simulation environments and synthetic data generation.
• Strong understanding of multi-modal sensor fusion (LiDAR, depth cameras, IMUs) and time-series modeling.
• Experience working with different AI models, including LLMs, generative AI, diffusion models, and reinforcement learning algorithms.
• Programming: Proficiency in Python, C++, and experience with robotics frameworks (ROS2, PyTorch, TensorFlow, JAX, OpenCV).ML Techniques: Deep learning (CNNs, Transformers), reinforcement learning (PPO, SAC, DDPG), and self-supervised learning.
• Simulation & Modeling: Experience with physics-based simulation tools (Isaac Sim, Mujoco, Gazebo, PyBullet).Data Processing: Expertise in large-scale dataset curation, sensor fusion, and distributed ML pipelines.
• Optimization & Deployment: GPU acceleration, CUDA, TensorRT, and real-time model optimization.
• Experience with edge AI and deploying ML models on embedded robotics platforms.
• Background in multi-agent reinforcement learning (MARL) for robotic coordination.
• Understanding of generative AI (Diffusion Models, Transformers) applied to robotic environments.
• Experience with foundation models and fine-tuning large-scale pretrained ML models for robotics.
• Passionate about AI for robotics and solving hard, real-world problems.
• Strong R&D mindset, excited about prototyping and pushing AI innovation.
• Self-driven, curious, and thrives in an experimental, fast-paced environment.
• Effective collaborator, able to work cross-functionally with robotics and software teams.
• Willing to work onsite at our Cambridge,MA office atleast 3 days a week 

About Pickle RobotPickle Robot is a pioneer in Physical AI for supply chain applications. Today Pickle robots autonomously unload trucks, trailers, and import containers at human-scale or better performance. The alternative is manual work that is difficult, dirty, sometimes dangerous, and increasingly hard to staff at distribution centers around the globe. Pickle Robot is laser focused on automating truck unloading using generative AI, machine learning, computer vision, advanced sensors, and industrial robotics to deliver engineered products customers rely on. Pickle Robot Unload Systems work alongside people on loading docks to make the work safer, faster, and more efficient. Pickle robots are physical AI that unload trucks.
Pickle provides best-in-class benefits including health, dental, & vision insurance; unlimited vacation, along with all federal and state holidays; 401K contributions of 5% your salary, travel supplies, and other items to make your working life more fun, comfortable, and productive.