Machine Learning Engineer, Driving Foundation Models

What you'll do

• Curate and manage large datasets, including filtering, deduplication, tokenization, and synthetic data generation.
• Train models across GPU/TPU clusters; implement post-training alignment and optimize inference for cloud and edge hardware.
• Build evaluation benchmarks and safety guardrails, including hallucination checks and red-teaming pipelines.

What they ask for

• Programming languages: Python, C++, and Golang.
• Experience with PyTorch, JAX, the Hugging Face ecosystem, and vLLM.
• Knowledge of Kubernetes, Docker, distributed training, and cloud platforms.
• Knowledge of linear algebra, calculus, probability, statistics, and deep learning architectures.

Tools

Python, C++, Golang, PyTorch, JAX, Hugging Face, vLLM, Kubernetes

AI focus

Develop driving and physical AI foundation models, support Aria VLA model development, and perform data curation, distributed training, alignment, inference optimization, and safety evaluation.