Principal Scientist, ML – Overall Architect
Core
Principal-level architect owning the end-to-end technical coherence of Adobe's next-generation video and image foundation models, bridging distributed training systems, inference infrastructure, and data pipelines.
Role type
Principal Scientist, ML – Overall Architect
Builds
Next-generation generative video and image foundation models (GenRender6, Gen6.5, GenEdit1)
Domain
Generative AI, Computer Vision, Large-scale ML Systems
Deliverable
production ML models
Required skills
Distributed training systems expertise, Inference and serving architecture, Model architecture design, Data pipeline design, Cross-team technical leadership, Scalability optimization
Preferred skills
Generative AI research publications, End-to-end ownership of video/image foundation models, Cost-aware architecture decisions
Technologies
PyTorch, FSDP, Tensor Parallelism, Pipeline Parallelism
Responsibilities
Own overall technical architecture spanning training, inference, and data; Provide technical direction for large-scale distributed training and inference systems; Drive co-design between model architecture and data; Serve as primary technical authority connecting teams; Own architecture decisions for specific generative models; Optimize architecture for performance, quality, and cost.
Seniority
Principal, hands-on IC with strategic scope