Robotics Engineer, Manipulation — Real-World Deployment

Frontdoor Defense; Sunnyvale (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Apptronik in Sunnyvale, CA seeks a Software Engineer - Manipulation to deploy learning-based policies from simulation to our robot fleet. You will bridge research and production software, delivering reliable, real-time manipulation capabilities on physical hardware.

You will collaborate with simulation, data, hardware, and systems teams to move policies from demonstrations through post-training to robust deployment, focusing on visuomotor control for grasping and whole-arm manipulation.

Qualifications

• Must have manipulation experience with hardware deployment.
• Experience training and debugging learned policies including RL fine-tuning, distillation, and imitation learning (e.g., DAgger).
• Proficiency in Python, C++, and PyTorch or JAX with real-time robotic software experience.
• Experience with demonstration and teleoperation data and building evaluation pipelines for learned policies.

Responsibilities

• Implement and deploy state-of-the-art learning-based manipulation policies on real hardware.
• Adapt policies for new tasks and hardware using post-training methods.
• Build and maintain data pipelines for teleoperation, demonstration collection, and dataset curation.
• Drive development from simulation to robust real-world deployment on the robot fleet.
• Develop visuomotor policies for grasping and manipulation with low latency.

Skills

Manipulation

Learning-Based Control

Software Engineering

Data & Evaluation

Education

MS degree with 3+ years experience

PhD in related field

Tools

IsaacSim

MuJoCo

Drake

PyTorch

JAX

Python

C++

Job description

Apptronik in Sunnyvale, CA seeks a Software Engineer - Manipulation to deploy learning-based policies from simulation to our robot fleet. You will bridge research and production software, delivering reliable, real-time manipulation capabilities on physical hardware.

You will collaborate with simulation, data, hardware, and systems teams to move policies from demonstrations through post-training to robust deployment, focusing on visuomotor control for grasping and whole-arm manipulation.