Multimodal Foundation Models Engineer - Scalable Driving AI

Deeproute.ai; Fremont (CA); In loco; USD 150.000 - 230.000; Tempo pieno

Descrizione del lavoro

Deeproute.ai in Fremont, CA seeks researchers and engineers to advance multimodal foundation models for autonomous driving. Bridge research and production with a focus on robustness, scalability, and real-world deployment in large-scale systems.

Ideal candidates bring hands-on engineering, strong experimental rigor, and expertise in self-supervised and representation learning, with experience in vision-language-action modeling and scalable pretraining.

Competenze

• Ideal candidates have strong experimental rigour and ability to bridge research and production.
• Hands-on engineering experience and a systems mindset are highly valued.
• Experience with vision-language-action modeling, multimodal learning, and large-scale pretraining.
• Publication record in top venues is preferred.

Mansioni

• Large-Scale Foundation Model Pretraining: develop scalable preprocessing and training pipelines.
• Work on vision-language-action models, video foundation models, and long-context temporal modeling.
• Improve training stability, data efficiency, scaling, and representation robustness.
• Develop distributed training systems using PyTorch Distributed, DeepSpeed, and Megatron-LM.

Conoscenze

Strong experimental rigor

Solid systems and modeling intuition

Hands-on engineering ability

Interest in scalable multimodal AI

Foundation models

Self-supervised learning

Representation learning

Multimodal learning

Large-scale pretraining

Strong publication record

Formazione

MS or PhD in Computer Vision, ML, Robotics, CS

Strumenti

PyTorch Distributed

DeepSpeed

Megatron-LM

Descrizione del lavoro

Deeproute.ai in Fremont, CA seeks researchers and engineers to advance multimodal foundation models for autonomous driving. Bridge research and production with a focus on robustness, scalability, and real-world deployment in large-scale systems.

Ideal candidates bring hands-on engineering, strong experimental rigor, and expertise in self-supervised and representation learning, with experience in vision-language-action modeling and scalable pretraining.