Engineer Agentic AI - Inventory Solutions; Hybrid - Seattle, WA)

Nordstrom runs on knowing what inventory it has, where it is, and whether the record can be trusted. The Inventory Integration Systems team owns that truth across a multi-market, multi-channel business — the platforms that track in-transit inventory, reconcile audits, capture vendor data, and keep enterprise inventory accurate for every downstream system that depends on it.

We are now applying agentic AI to that mission. Across platforms like AURA (audit unification using RAG-based AI), IAP (Inventory Audit Platform), Omni Trace (in-transit dashboard and ledger), and our PADLC agent pipeline, we're building agents that don't just report on inventory problems — they investigate them, correlate signals, and take scoped action. This is applied AI on a real, high-stakes retail domain, not a research sandbox.

• Implement individual components of agentic inventory solutions — tool integrations (Jira, Slack, Confluence, internal inventory APIs), prompt templates, and retrieval steps — under the direction of senior engineers, writing clean, well-tested code and documenting your work clearly.
• Learn how the team applies context engineering — what an agent sees, when, and why — including how we ground agents in inventory context (audit records, ledger state, vendor data) so their answers are accurate and traceable.
• Help build and run evaluation scripts and offline benchmarks to check that agent behavior is safe, accurate, and reliable — especially where inventory decisions carry financial implications.
• Participate in on-call rotations and design reviews, learning to debug issues and surface trade-offs on the specific functions or components you own.
• Add basic instrumentation — logging, metrics, alerts — to the systems you build, and follow Nordstrom's engineering and security standards, including secure handling of sensitive inventory and vendor data.
• Collaborate across engineering, product, and business stakeholders to understand how your work fits into the larger inventory platform and where agentic AI can meaningfully reduce manual audit and reconciliation effort.

• 1+ years of professional software engineering experience (internships count), or a relevant certification (e.g. bootcamp).
• AI Fluency Required — Some hands-on exposure to LLMs or foundation model APIs (Anthropic, OpenAI, Google, etc.), whether through coursework, projects, internships, or professional experience.
• Genuine curiosity about agentic AI — how agents use tools, retrieve context, and take actions — and a desire to build production skills in this space.
• Solid CS fundamentals — data structures, algorithms, and object-oriented design — along with familiarity with RESTful APIs, JSON, and how services communicate with one another.
• Basic understanding of software testing practices and version control (e.g. Git).
• Willingness to learn modern agentic frameworks such as Lang Graph, CrewAI, Semantic Kernel, or the Claude Agent SDK.
• Strong verbal and written communication skills, and a collaborative, feedback-seeking mindset.

• Coursework, personal projects, or internship experience involving prompt engineering, retrieval-augmented generation (RAG), or vector databases.
• Exposure to a backend language and framework such as Java / Spring Boot or Python, and to messaging middleware like Kafka.
• Exposure to cloud platforms such as AWS or GCP, containerization (Docker / Kubernetes), and CI/CD basics.
• Familiarity with backend databases (e.g. Spanner, Postgres).
• Interest or coursework in retail, e-commerce, supply chain, or inventory problem spaces.
• Any contributions to open-source AI projects or personal experimentation with LLM applications.

T…