The OpportunityAdobe Applied Science & Machine Learning (ASML) is seeking a
Principal Scientist, ML - Overall Architect to serve as the cross-cutting technical owner bridging our training and inference framework, ML model building, and data across Adobe's next-generation video and image foundation models.
In this role, you will operate as the
principal-level architect responsible for end-to-end coherence - holding the full technical picture across large-scale distributed training systems, inference and deployment, model architecture, and data recipe design, and making the architecture decisions that keep all of them aligned. Rather than owning a single system or model, your scope spans the seam between infrastructure, modeling, and data, ensuring these interdependencies are resolved at the architecture level rather than through coordination overhead.
This role is intended for those who can operate with genuine authority at the intersection of systems, modeling, and data at scale - and whose technical judgment drives cross-team direction.
Job ResponsibilitiesEnd-to-End Architecture Ownership Own the overall technical architecture spanning the Training & Inference Framework, ML model building, and data - making principled design decisions that keep training systems, model architecture, and data pipelines coherent and mutually reinforcing.
Training & Inference Systems Leadership Provide principal-level technical direction for large-scale distributed training (FSDP, Tensor Parallelism, Pipeline Parallelism, checkpointing, fault tolerance) and inference/serving infrastructure, ensuring systems are performant, reliable, and cost-efficient at scale.
Model Architecture & Data Co-Design Drive the interaction between model architecture choices, training recipes, and data design - identifying co-design opportunities and resolving architectural tensions across modeling and data teams.
Cross-Team Technical Direction Serve as the primary technical authority connecting TIF, model building, and data teams - unblocking architectural seams, aligning on interfaces, and ensuring workstreams remain coherent as models scale.
Architecture Decisions for Next-Generation Models Own architecture-level decisions for GenRender6 / Gen6.5 and GenEdit1, spanning training execution, inference deployment, model design, and data pipeline integration.
Performance and Scalability Drive architecture choices that optimize for training and inference efficiency, model quality, and cost - translating scaling and deployment requirements into concrete architecture decisions.
What You'll Need to Succeed- Education: PhD in Computer Science, Electrical Engineering, AI/ML, or a related field, or equivalent depth demonstrated through research contributions or principal-level systems impact.
- Principal-Level Technical Breadth: Demonstrated ability to reason authoritatively across distributed training systems, inference/deployment infrastructure, model architecture, and data pipelines - and to make architecture decisions that span all of them.
- Deep Distributed Training & Inference Expertise: Hands-on expertise with large-scale distributed training (PyTorch FSDP, tensor and pipeline parallelism, checkpointing, fault tolerance) and inference/serving for large generative models.
- Model Architecture & Data Understanding: Strong understanding of how model architecture decisions interact with training infrastructure and data recipes, with the ability to drive principled co-design across these areas.
- Cross-Team Architecture Leadership: Proven track record of driving technical direction across multiple teams, resolving architectural conflicts, and delivering systems at the principal/staff level in a complex organization.
- Senior-Level Judgment & Execution: Ability to cut through ambiguity, identify the highest-leverage architectural decisions, and drive them to completion with organizational credibility and technical rigor.
Preferred Experience- Research contributions or publication record in generative AI, computer vision, or large-scale ML systems.
- Experience owning end-to-end architecture for video, image, or multimodal foundation models from training through production deployment.
- Prior work as a technical anchor spanning systems and modeling teams in a large generative AI organization.
- Track record of shipping large-scale generative models (video generation, diffusion, flow matching) to production.
- Experience with cost-aware architecture decisions for large model deployments.
Expected Pay Range:Our compensation reflects the cost of labor across several U.S. geographic markets, and we pay differently based on those defined markets. The U.S. pay range for this position is $206,400 -- $379,100 annually. Pay within this range varies by work location and may also depend on job-related knowledge, skills, and experience. Your recruiter can share more about the specific salary range for the job location during the hiring process.
In California, the pay range for this position is $261,800 - $379,100
At Adobe, for sales roles starting salaries are expressed as total target compensation (TTC = base + commission), and short-term incentives are in the form of sales commission plans. Non-sales roles starting salaries are expressed as base salary and short-term incentives are in the form of the Annual Incentive Plan (AIP).
In addition, certain roles may be eligible for long-term incentives in the form of a new hire equity award.
