Autonomy Engineer - VLA Pre-training

$180,000 – $300,000 · Posted 1 day ago

San DiegoLast seen today

Summary

As an Autonomy Engineer focused on VLA (Vision-Language-Action) pre-training, you will develop and train deep learning policies for humanoid robots, owning the full pipeline from data collection and curation to model deployment. You'll pre-train base models on diverse multi-embodiment trajectory corpora, fine-tune policies for specific tasks, work with MLOps teams to scale distributed training, and build continuous pipelines that ingest synthetic data and teleop logs. This is primarily a deep learning role requiring 3+ years of experience shipping neural network models, hands-on expertise with LLMs, VLMs, or generative models, and strong proficiency in Python and PyTorch or JAX.

Data Science Machine Learning Python AI Behavior Cloning Checkpointing Data Curation Deep Learning Distributed Training Edge Inference Jax LLM Large Language Models Mlops Model Inference Model Profiling Openvla Physical Intelligence PyTorch Reinforcement Learning Rl Streaming Datasets Synthetic Data Vision Language Action Vla Vlm Weak Supervision