Über diese Software Engineer V – Applied ML Research Stelle bei Picklerobot
About Us
At Pickle Robot, we're on a mission to automate global supply chains with Physical AI. Our robots work alongside warehouse teams to unload trucks and containers — one of the toughest, most understaffed jobs in logistics — making the work safer, faster, and more efficient for the people doing it. Loading trucks comes next, followed by the Dill Autonomy Engine: generalized autonomy that will eventually orchestrate robots across entire logistics processes.
We're looking for an applied ML engineer to own one of our hardest technical problems: how a robot decides what to grasp, and how to grasp it. The answer is a fusion of classical robotics with state-of-the-art neural networks, and you'd be the person setting its direction on a team of two.
This is a senior seat — you'll be trusted to pick the approach, not just implement one. If you've gotten your hands dirty with both ML systems and real robot platforms, and you want problems that are technically deep and practically grounded, this is the role.
What You'll Do
Spearhead Robotic Grasping: Take end-to-end ownership of the intelligence driving our robots' physical interactions, dynamically determining optimal gripper configurations and tool selections for diverse, complex objects.
Bridge Research and Reality: Translate cutting-edge machine learning research into robust production systems. You will modify novel architectures, orchestrate training pipelines, deploy directly to our robotic fleet, and drive continuous improvement based on real-world field telemetry.
Architect Ground-Truth Evaluation: Design and implement rigorous evaluation infrastructure that proves model efficacy in live, chaotic environments (like real trailers) — moving beyond static benchmarks to guarantee true operational performance.
Pioneer Simulation-First Development: Engineer and validate sophisticated grasping and manipulation strategies within high-fidelity simulations to ensure flawless execution before deploying to physical hardware.
Drive Cross-Functional Technical Vision: Collaborate closely with perception and motion planning experts, setting the technical bar through rigorous code reviews, innovative design discussions, and collaborative debugging.
Leverage Force-Multiplying Workflows: Utilize advanced AI coding assistants and modern development paradigms to accelerate engineering velocity, allowing you to achieve outsized impact and move faster than traditional teams.
What You'll Bring
Proven ML Track Record: 6+ years of experience shipping end-to-end machine learning systems — spanning data pipelines, training, evaluation, and deployment — and maintaining ownership in production. (An MS degree counts as two years of experience).
Applied Robotics Expertise: Demonstrated success deploying machine learning models within complex, active physical robotic systems.
Architectural Depth: Deep expertise in bespoke model design. You don't just rely on off-the-shelf APIs; you possess the ability to fundamentally reshape neural networks to solve unique, domain-specific challenges.
Cloud ML Proficiency: Fluency and comfort in architecting and running medium-to-large-scale training pipelines in cloud environments.
Engineering Excellence: Mastery of Python and modern software engineering best practices.
Foundational Robotics Knowledge: A solid grasp of classical robotics concepts, including motion planning, kinematics, geometry, and perception.
Academic Foundation: A BS or higher in Robotics, Computer Science, or a closely related technical field.
Cross-Disciplinary Communication: The ability to distill complex ML concepts and collaborate effectively with engineers across different specialized domains.
Local Collaboration: Ability to work from our dynamic Charlestown, MA headquarters at least 3 days a week in a hybrid capacity.
Preferred
A background spanning both perception and motion planning.
Simulation tooling for robotics development (Isaac Sim, MuJoCo).
GCP specifically.
AI coding assistants already part of your daily workflow.
