Multimodal AI Engineer — Image/Video Synthesis

SpaceX Super Intelligence; Seattle (WA); Sur place; USD 180 000 - 440 000; Plein temps

Avantages offerts par ce poste

Equity

Medical

Dental

Vision

401(k)

Disability Insurance

Life Insurance

Résumé du poste

SpaceXSI is seeking a multimodal engineer on the Imagine Model Team to advance image and video understanding, generation, and audio integration for richer experiences. You will work across data curation, modeling, training, and production integration, spanning pretraining and post-training phases.

You will collaborate with product teams to push model frontiers and deliver end-to-end user experiences, driving high-quality visual content capabilities across platforms.

Qualifications

• Track record leading studies that significantly improve neural network capabilities through data or modeling.
• Experience in data-driven experiment designs, systematic analysis, and iterative model debugging.
• Experience developing or working with large-scale distributed machine learning systems.

Responsabilités

• Drive multimodal capabilities focusing on image and video generation, editing, and understanding.
• Improve data quality through annotation, filtering, augmentation, and synthetic data generation.
• Design evaluation frameworks, metrics, benchmarks, and reward models for quality and coherence.
• Implement efficient algorithms for real-time inference, distillation, and scalable serving.
• Develop data collection and processing pipelines for multimodal datasets.
• Collaborate cross-functionally to integrate AI solutions into production and iterate from user feedback.

Connaissances

Multimodal ML

Distributed ML

Data quality

Python

Model debugging

Cross-functional collaboration

RL training

Video/image processing

Outils

PyTorch

JAX

Rust/C++

Spark

Ray

Description du poste

SpaceXSI is seeking a multimodal engineer on the Imagine Model Team to advance image and video understanding, generation, and audio integration for richer experiences. You will work across data curation, modeling, training, and production integration, spanning pretraining and post-training phases.

You will collaborate with product teams to push model frontiers and deliver end-to-end user experiences, driving high-quality visual content capabilities across platforms.