below marketmost similar roles pay hereabove market
This listing doesn't post a salary. Most similar roles pay
$195,150–$254,750.
Based on 240 similar postings.
Employer
About Adobe
Adobe Inc. is a global software company known for creative and multimedia software products including Photoshop, Illustrator, Acrobat, and its cloud-based Creative Cloud and Document Cloud suites. Industry: Creative & Digital Experience Software
Adobe currently has
282 open roles
on FindRole.
Listed pay typically runs
$148,100–$282,100
across 76 roles with salary data.
The Principal Scientist, ML – Overall Architect joins the Applied Science & Machine Learning team to serve as the technical owner bridging training and inference frameworks, model building, and data for next-generation video and image foundation models. This role involves making high-level architecture decisions to ensure end-to-end coherence across large-scale distributed training systems, inference deployment, and data recipe design. You will drive the interaction between model architecture choices and training recipes while providing technical direction for distributed training using PyTorch FSDP, tensor parallelism, pipeline parallelism, checkpointing, and fault tolerance. The position requires expertise in generative AI, computer vision, and large-scale ML systems to solve complex scaling and deployment requirements for models like GenRender6 and GenEdit1, ensuring performance, reliability, and cost-efficiency across the entire production lifecycle.
Own the end-to-end technical architecture for training frameworks, model building, and data pipelines.
Provide technical direction for large-scale distributed training systems including FSDP and pipeline parallelism.
Drive the co-design of model architectures, training recipes, and data design to resolve technical tensions.
Serve as the primary technical authority to align interfaces and unblock architectural seams across multiple teams.
Make architecture-level decisions for next-generation video and image foundation models.
Optimize architecture choices for training and inference efficiency, model quality, and cost-effectiveness.
What we're looking for
PhD in Computer Science, Electrical Engineering, AI/ML, or a related field, or equivalent depth demonstrated through research or principal-level systems impact.
Demonstrated ability to reason authoritatively across distributed training systems, inference/deployment infrastructure, model architecture, and data pipelines.
Hands-on expertise with large-scale distributed training including PyTorch FSDP, tensor and pipeline parallelism, checkpointing, and fault tolerance.
Experience with inference/serving for large generative models.
Strong understanding of how model architecture decisions interact with training infrastructure and data recipes to drive principled co-design.
Proven track record of driving technical direction across multiple teams and resolving architectural conflicts at the principal or staff level.
Research contributions or publication record in generative AI, computer vision, or large-scale ML systems (preferred).
Experience owning end-to-end architecture for video, image, or multimodal foundation models from training through production deployment (preferred).
Python
PyTorch
Generative AI
Diffusion Models
VLM
Distributed Training
Multi-modality
Deep Learning
Data Curation
Captioning Pipelines
Image-to-Image
Mid-training
Pre-training
Post-training
Generative AI
NLP
Computer Vision
Deep Learning
LLM
PyTorch
TensorFlow
Keras
JAX
Hugging Face
CUDA
Python
MLOps
LLMOps
RAG
GraphRAG
Vector Databases
LoRA
QLoRA
AWS
GCP
Azure
Distributed Training
Model Serving
Feature Stores
A/B Testing
Generative AI
PyTorch
JAX
vLLM
SGLang
Large Language Models (LLMs)
Distributed Training
Inference Optimization
Transformer Architectures
MuJoCo
Deep Learning
Parallelism Strategies
KV-caching
Speculative Decoding
Low-precision Training
Foundation Models