42dot
Machine Learning Engineer, Driving Foundation Models
Sunnyvale, United States. Posted 16 days ago
The role
Develop driving and physical AI foundation models. Support Aria VLA model development.
What you will do
- Design and manage massive, multi-terabyte datasets including scraping, filtering, deduplication, tokenization, and synthetic data generation.
- Scale training across massive GPU/TPU clusters using parallel processing, custom data loaders, and framework optimization.
- Implement instruction tuning, Reinforcement Learning from Human Feedback (RLHF), or direct preference optimization.
- Compress and optimize models to run at low latency on cloud infrastructure or edge hardware.
- Build benchmarking harnesses, hallucination checkers, and red-teaming pipelines to test model accuracy and safety.
Required qualifications
- Python
- C++
- Golang
- PyTorch
- JAX
- Hugging Face ecosystem
- vLLM
- Kubernetes
- Docker
- multi-node distributed training (Megatron-LM, DeepSpeed)
- cloud platforms (AWS, Azure, Google Cloud)
- Linear algebra, calculus, probability, statistics, and deep learning architectures (Transformers, diffusion models)
About the company
42dot is a mobility AI company and the Global Software Center of Hyundai Motor Group. They develop software-defined vehicle technologies and advance software and AI technology.