Want to know if this job is worth applying to?
San Francisco, California, United States
Remote only
Own the workflows within the end-to-end orchestration layer that connects our product-to-market capabilities; defining the workflows from one product to the next, where agents act autonomously, where humans decide, and how context, state, and intent are passed between agents, systems, and users.
Translate future-state workflow blueprints into executable orchestration: the triggers, conditions, handoffs, exception paths, and human-in-the-loop checkpoints our platform operates against day to day.
Partner with capability product managers across design, merchandising, planning, and buying to ensure each product exposes the inputs, outputs, and signals required for cross-workflow orchestration; turning a portfolio of capabilities into a coherent operating system.
Design human-AI collaboration patterns that calibrate trust over time: where agents recommend, where they act, where they request confirmation, and how decision rationale, confidence, and provenance are surfaced to users.
Define agent evaluation and observability requirements; how we measure whether agents are doing the right thing, how we detect drift or regression, and how we close the loop between user feedback, eval data, and orchestration logic.
Establish the orchestration data model: the shared vocabulary of entities, states, events, and decisions every capability and every agent operates against, in partnership with Architecture, Data, and Engineering.
Lead the blueprinting and process design for cross-capability workflows, drawing on VOC, journey mapping, and process diagnostics to identify the friction points where orchestration creates the most leverage.
Drive change management for new ways of working; partnering with business stakeholders to shift teams from manual, sequential workflows to AI-orchestrated, parallelized ones, including training, documentation, and adoption measurement.
Establish and track OKRs tied to workflow cycle time, decision velocity, agent reliability, human intervention rate, and adoption across global teams.
Partner with the platform team on agent capabilities, tool integrations, memory, and policy controls required to run production workflows at retail scale.
Help evolve portfolio governance for the product-to-market ecosystem, ensuring orchestration priorities reflect both business value and the readiness of underlying capabilities and data.
5+ years in product management, product operations, or workflow/process design; ideally in retail, consumer product, or enterprise SaaS environments driving complex cross-functional transformation.
Hands-on familiarity with agentic AI: you've shipped or worked closely on products involving LLM-based agents, tool use, multi-step reasoning, retrieval, or human-in-the-loop systems, and you understand the design patterns (and failure modes) that come with them.
Strong process-design instincts: comfortable mapping complex workflows into blueprints, identifying where automation belongs vs. where human judgment must stay, and designing for graceful exception handling.
Deep understanding of product creation workflows across design, planning, buying, and allocation. Enough to know where the real handoffs, decisions, and data dependencies live.
Fluent in the language of agents: prompts, tools, memory, evaluation, guardrails, confidence, escalation. You don't need to train models, but you do need to design what they do and how they behave.
Track record building shared platforms or operating layers - not just point features - and partnering with engineering on data models, APIs, and event-driven architectures.
Skilled at driving clarity across diverse audiences: creative teams, planners, engineers, ML practitioners, and executives. You can hold a workflow blueprint and an agent architecture diagram in the same conversation.
Analytical, evidence-driven, and biased toward measurement: you instrument what you ship and use evaluation data to evolve orchestration logic over time.
User-centric thinker who prioritizes trust, transparency, and explainability in AI-augmented workflows. You design for the moment users decide whether to trust the agent.
Bachelor's degree in Business, Design Management, Engineering, Computer Science, or a related field; advanced degree or coursework in HCI, service design, or AI/ML a plus.