Senior Data Scientist – AI & ML | MLOps Enablement (Hybrid – Seattle, WA)

Nordstrom•Seattle, WA

2d•Hybrid

About The Position

The Developer Platform Organization’s mission is to accelerate the delivery of reliable and secure platforms that make developers feel good and code their best. Developer Platform exists to help engineers focus on business challenges and minimize their work on infrastructure and operations — developing and supporting platforms and tools for the entire Software Development Lifecycle. Centralized platform tooling allows developer tooling to be written once, and not repeated for each team or project. Within Developer Platform, the MLOps Enablement team owns the ML Platform capability. Data Scientists and engineers can build, deploy, and operate machine learning models on managed, standards-compliant infrastructure — without standing up their own model serving or ML pipeline tooling. We deliver a unified, secure, and cost-efficient platform built on Vertex AI. We are looking for a Senior Data Scientist to join the MLOps Enablement team as an embedded DS practitioner. This is not a traditional Data Science role focused on owning models — it is a platform-facing role for a DS practitioner who wants to shape the infrastructure and tooling that Data Scientists across Nordstrom depend on every day. You will be the DS voice on a platform engineering team, ensuring our capabilities are designed for how Data Scientists actually work — so adoption is fast, intuitive, and does not require a custom engagement every time. This role is offered as hybrid in Seattle, WA. Candidates must be available to work in office at the Nordstrom corporate headquarters a minimum of 4 days/week to be considered for this position.

Requirements

Bachelor’s, Master’s, or PhD in Statistics, Data Science, Computer Science, Engineering, or a related technical field required
10+ years of hands-on Data Science experience with production model delivery across multiple ML (classification, ranking, NLP, time-series, recommendation) and GenAI models
Deep expertise in model evaluation — defining metrics, thresholds, and evaluation pipelines for real-world production models
Experience with Feature Store design, feature engineering, and understanding of feature freshness, reuse, and drift across different model families
Proficiency in Python with experience writing clean, maintainable, production-quality ML code
Strong understanding of ML monitoring — data drift, prediction drift, and concept drift detection
Experience with experiment tracking and model lifecycle management
Ability to translate between DS practice and platform engineering — comfortable driving design decisions, authoring DS-native documentation, and engaging in technical design reviews
Self-directed; comfortable owning POC work end-to-end without a dedicated DS team structure

Nice To Haves

Hands-on experience with GCP and Vertex AI — Workbench, Pipelines, Feature Store, Model Endpoints, Model Registry, Model Evaluation (preferred)
Familiarity with AWS SageMaker for cross-cloud benchmarking and comparison context (preferred)
Understanding of CI/CD for ML, containerization, and pipeline orchestration — able to engage at platform depth alongside MLOps engineers (preferred)
Prior experience in ML platform adoption, enablement, or developer experience work (preferred)
Experience operating within a mature ML lifecycle — versioning, lineage tracking, model governance, staged rollouts, and model deprecation practices at enterprise scale (preferred)
Exposure to agentic AI patterns, LLM evaluation frameworks, or Vertex AI Agent Builder (preferred)

Responsibilities

Run end-to-end POC validation for new platform capabilities — Feature Store, Endpoints, Model Evaluation, AutoML, BigQuery ML etc. — independently, before they reach DS teams at scale
Attend DS team planning and design sessions as an embedded practitioner; surface real workflow pain points and translate them into reusable MLOps platform requirements
Design and own the Model Evaluation Framework — defining metrics, thresholds, and evaluation pipelines for batch, online, and streaming use cases on Vertex AI
Build model-type-aware Feature Store schemas, endpoint configurations, and evaluation pipelines that accommodate the fundamentally different needs of different ML models
Lead benchmarking of Nordstrom’s platform against industry standards — SageMaker vs. Vertex AI — across feature parity, cost, and DS practitioner ergonomics
Author DS-native documentation, onboarding guides, and quickstart notebooks that lower the adoption barrier for new platform features
Contribute DS domain expertise to the emerging Vertex AI Agentic Platform — identifying DS workflow pain points as agent use cases and defining evaluation frameworks for agentic responses
Own model card standards — capturing what actually matters to a practitioner, not just governance checkboxes
Communicate complex trade-offs and platform decisions to technical and non-technical stakeholders across DS, engineering, and leadership