Apptronik
MLOps Engineer
About this role
Apptronik seeks a Staff MLOps Engineer to architect and lead their ML platform, owning the end-to-end model lifecycle from data collection and teleoperation to deployment on their humanoid robot Apollo. You'll design the dataset versioning, experiment tracking, model registry, evaluation harnesses, and serving infrastructure that connects research to production robotics at scale.
What you'll do
- Define MLOps platform architecture, subsystem interfaces, and engineering standards for dataset and model workflows across teams
- Design and operate dataset versioning, lineage tracking, and reproducibility systems to ensure models are traceable to source data
- Build and maintain a model registry with versioning, metadata, evaluation results, and promotion workflows from training to deployment
- Create automated evaluation and qualification harnesses that gate models before deployment to Apollo robots
- Own the serving and packaging pipeline for deploying models to robots, including ONNX/TensorRT optimization and on-device versioning
- Mentor MLOps engineers and establish cross-functional standards with Autonomy, Data Platform, and TeleOp teams
What they're looking for
- Python and systems-level languages (Go, Rust, or C++)
- MLOps platform design and production ML infrastructure
- Dataset versioning tools (DVC, LakeFS, Delta, or similar)
- Experiment tracking systems (MLflow, W&B, Determined)
- Model registry and artifact management
- Kubernetes and service-oriented architecture
- Policy serving and model optimization (ONNX, TensorRT, torch.compile)
- Robotics or autonomous systems experience (preferred)
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Apptronik
Apptronik builds Apollo, a humanoid robot designed for manufacturing, logistics, and other industrial applications. The company is hiring firmware engineers, software engineers, controls specialists, and networking engineers to develop embedded systems, motor control, motion pipelines, and wireless connectivity that power the robot's dexterous manipulation and autonomous operation.
- Website
- apptronik.com
Likely interview questions
- Walk us through an end-to-end MLOps platform you've built or significantly contributed to—how did you structure the dataset, experiment, model registry, and serving components?
- How have you handled reproducibility and lineage tracking across datasets, training code, and model artifacts in a production environment?