Senior Multimodal AI Engineer

SpaceX Super Intelligence; Palo Alto (CA); On-site; USD 180,000 - 440,000; Full time

Job summary

SpaceXSI in the United States is building a multimodal AI stack and seeks a leading engineer to advance superhuman multimodal intelligence across image, video, audio and text. You’ll work on data curation, tokenizer training, large-scale pre-training, alignment, scaling, evaluation, tooling, and end-to-end product experiences.

Collaborate with cross-functional teams to deliver frontier capabilities in multimodal reasoning, world modeling, tool use, and real-time human-AI collaboration.

Qualifications

• Experience with multimodal pre-training, post-training, or fine-tuning.
• Proficiency in Python; experience with JAX/PyTorch/XLA.
• Experience building large-scale distributed ML systems.
• Designing data pipelines at scale for noisy multimodal data.
• Knowledge of evaluation design, benchmarks, or RL techniques.
• Self-starter in high-intensity environments; willing to own initiatives.

Responsibilities

• Design, build, and optimize large-scale distributed systems for multimodal pre-training and inference.
• Develop high-throughput data acquisition, preprocessing, and management pipelines.
• Advance multimodal capabilities across modalities and real-time processing.
• Drive data quality studies and scalable pipelines for trillion-parameter models.
• Create evaluation frameworks, benchmarks, and reward models.
• Innovate on algorithms, co-design hardware/software, and scaling paradigms.
• Build research tooling, demos, and end-to-end product experiences.
• Collaborate across stack from pre-training to post-training for real-time interaction.

Skills

Multimodal pre-training

Python

PyTorch

JAX

Distributed ML

Data pipelines

RL techniques

Self-starter

End-to-end ownership

Tools

Rust

C++

Spark

Ray

Kubernetes

Job description

SpaceXSI in the United States is building a multimodal AI stack and seeks a leading engineer to advance superhuman multimodal intelligence across image, video, audio and text. You’ll work on data curation, tokenizer training, large-scale pre-training, alignment, scaling, evaluation, tooling, and end-to-end product experiences.

Collaborate with cross-functional teams to deliver frontier capabilities in multimodal reasoning, world modeling, tool use, and real-time human-AI collaboration.