Research Engineer
Brak informacji o wynagrodzeniu
SeniorFull-time
#434706·Dodano miesiąc temu·1
Źródło: NomagicTech Stack / Keywords
PythonPyTorchJAXTransformersImitation LearningReinforcement LearningVLMVLARoboticsMachine Learning
Firma i stanowisko
Nomagic is focused on developing general-purpose physical AI by leveraging massive deployment logs from real robotic systems operating in production environments. The company combines world-class machine learning research with industrial-scale robotic execution and has an established presence in this field.
Wymagania
- Deep experience at the intersection of machine learning, systems engineering, and robotics.
- Proven experience training, fine-tuning, and deploying deep learning architectures (e.g., Transformers, VLMs/VLAs, Imitation Learning, Reinforcement Learning) for robot control, ideally validated on real hardware.
- Strong proficiency in Python and deep learning frameworks such as PyTorch and JAX.
- Hands-on comfort with hardware and understanding of the robotics stack including perception, controls, and state estimation.
- Ability to move seamlessly between research and implementation with emphasis on execution, iteration speed, and real-world robustness.
Obowiązki
- Focus on the intersection of Robotics and Machine Learning with large-scale multimodal model training.
Core Research & Large-Scale Infrastructure:
- Design, implement, and maintain infrastructure for large-scale VLA model training, including scheduling, distribution, job management, checkpointing, and logging.
- Build tools for launching, monitoring, debugging, and reproducing complex experiments.
- Use offline classical stack data to pre-train robust robot foundation models.
- Translate research needs into infrastructure capabilities and analyze results to support ML researchers.
Real-World Evaluation & Operations:
- Design robotic tasks and lightweight physical setups to evaluate model capabilities beyond simulation.
- Coordinate data collection and execute structured on-robot evaluations to measure success rates.
- Analyze real-world results to guide ML research and identify operational bottlenecks.
- Beta test tools for teaching robots new skills and document workflows for team scalability.
Benefity
- Opportunity to work daily with real robots solving real problems.
- Relocation package.
- Flexible working hours.
- English-speaking environment.
Pakiet relokacyjny
Elastyczne godziny
Nomagic
4 aktywne oferty