STAFF, DATA SCIENTIST

Position Summary...

Drives the execution of multiple business plans and projects by identifying customer and operational needs; developing and communicating business plans and priorities; removing barriers and obstacles that impact performance; providing resources; identifying performance standards; measuring progress and adjusting performance accordingly; developing contingency plans; and demonstrating adaptability and supporting continuous learning.

Provides supervision and development opportunities for associates by selecting and training; mentoring; assigning duties; building a team-based work environment; establishing performance expectations and conducting regular performance evaluations; providing recognition and rewards; coaching for success and improvement; and promoting a belonging mindset in the workplace.

Promotes and supports company policies, procedures, mission, values, and standards of ethics and integrity by training and providing direction to others in their use and application; ensuring compliance with them; and utilizing and supporting the Open Door Policy.

Ensures business needs are being met by evaluating the ongoing effectiveness of current plans, programs, and initiatives; consulting with business partners, managers, co-workers, or other key stakeholders; soliciting, evaluating, and applying suggestions for improving efficiency and cost-effectiveness; and participating in and supporting community outreach events.

What you'll do...

About the role Merchandising Data Science builds the systems — both automated decision engines and decision support systems — that drive high-stakes merchandising decisions: how we select and curate items, manage space, plan and allocate inventory, flow inventory through the network, decide what and how much to buy, read the competitive landscape, and price.

As a Staff Data Scientist, you own a single, hard problem end-to-end — not an abstract domain, but a specific question like "how do we estimate price elasticities better?", "how do we reduce forecast error for seasonal items a year out?", or "how do we incorporate all constraints at DC outbound so the planned flow can actually be executed, exceptions and all?" Depending on the problem, the answer is a system that decides automatically, or one that gives a merchant or planner the right recommendation, tradeoff, and evidence to decide well.

You take the problem from a rough business question to a precise formulation to a production system, and you make sure it keeps getting better even on the days you're not in the room.

You are an individual contributor with no direct reports — your leverage comes from raising the judgment and standards of the people working alongside you on that problem, not from personal output alone.

What you'll do Own one hard, well-defined problem end-to-end — e.g., "how do we estimate price elasticities better?" or "how do we reduce forecast error for seasonal items a year out?" — from rough business question to precise formulation to production system.

Define what a good answer looks like: decision variables or recommendation logic, objective function, constraints, fallback behavior, and measurable success criteria.

Decide whether the problem calls for a fully automated decision engine or a decision support system that augments a human decision-maker, and design accordingly.

Choose and implement the right solution approach — optimization, heuristics, simulation, forecasting, causal/statistical inference, or a hybrid — and defend the choice on evidence.

Build the production system, not a prototype, integrating with the merchandising, planning, and platform systems it depends on.

Establish patterns, methods, and quality standards on this problem that other people working on it (Senior DS, engineers) adopt and extend.

Define the metrics and run the experiments that prove the answer is actually better — this may be estimation accuracy, forecast error, executability of a plan, margin, sell-through, or trust/adoption of a recommendation, depending on the problem.

Create mechanisms — documentation, reusable components, decision frameworks — so continued progress on the problem does not depend on you personally.

Proactively surface the next-most-important version of the problem before being asked (e.g., once elasticity estimation improves for core items, is the next gap seasonal or new items?).

Use AI-accelerated development (copilots, agents, evaluators) to speed iteration while holding a high bar for correctness and maintainability.

Communicate the problem, tradeoff, recommendation, and "so what" clearly to engineering leads, merchants, and business stakeholders.

What you'll bring A track record of taking a hard, specific business question — not a vague domain — and shipping a system, automated or decision-support, that measurably improved the answer.

Solid depth in optimization/decision methods (mathematical programming, constraint programming, heuristics, or simulation) and applied ML/forecasting/causal inference where relevant.

Experience with problems like: demand or elasticity estimation, long-horizon or seasonal forecasting, constrained planning/execution (e.g., DC outbound flow), assortment or allocation optimization, or competitive/price response modeling — in merchandising, retail, supply chain, or a comparable setting.

Comfort building for two different kinds of "users" — a machine executing a decision automatically, and a human (merchant, planner, analyst) who needs the right recommendation and evidence to decide well.

Demonstrated ability to make others working on the same problem better — through review, mentorship, or reusable standards — not just to produce more yourself.

Sound judgment on when to pursue optimality vs. a robust heuristic, and when to automate a decision vs. support a human making it, with the ability to explain the tradeoff.

A bias for iteration with accountability: you stay with a problem through adoption and validation, not just thro