Amazon Technologies, Inc.

Ali Sadighian

Deepak Bhatia

Chengliang Zhang

Tolga Han Seyhan

Salal Humair

Chao Liu

Jeffrey B Maurer

Jun Tong

Patrick Ludvig Bajari

Jingchen Wu

Jingqiao Zhang

Nicholas Deming Sherman

Di Wu

Zhihao Cen

Hernan P Awad

Andrew Bass

Abid Kapadya

Alvaro Maggiar

Julie Wei-Shu Chang

Stevie Robinson

Xiaochang Hu

Mahmoud Ahmed Bishr

Jialei Rong

Travis Francis Brayak

Haiyun Wang

Zhihao Chen

Techniques associated with determining a quantity of an item to inventory may be described. In particular, statistical data about variables associated with the item may be accessed. The statistical data may be generated based on historical realizations of the variables. The quantity may be computed based on this data. For example, a plurality of regrets corresponding to inventorying a plurality of quantities of the item may be generated based on the statistical data. The quantity may be determined from the plurality of quantities as one corresponding to an acceptable regret from the plurality of regrets.

Estimating an output based on robustness associated with multiple input variables

Embodiments herein are directed to determining an expected demand for an item. The specialized computer system can model an impact to an expected demand for the item between a current planning horizon and a next planning horizon, including an estimated leftover demand to be fulfilled during the next planning horizon. As calculated, this leftover demand may be less than an unfulfilled demand from the current planning horizon and may be estimated based at least in part on the unfulfilled demand and on historical leftover demand between planning horizons.

Inventory optimization based on leftover demand distribution function

Techniques for computing a feature based on variables may be provided. For example, historical realizations associated with a first variable may be accessed. The first variable may be associated with an item. Realizations of the first variable may be based on one or more factors associated with the item. Historical realization of a second variable associated with the item may also be accessed. The historical realizations of the first and second variables may be analyzed to generate an expected realization of the first variable as a function of the second variable. The feature may be computed based on the generated function.

Efficiently computing a feature based on a plurality of variables

Techniques are provided herein for utilizing an inventory engine to optimize the selection of a set of items to be stored as inventory at a storage location. A candidate set of items may be identified based at least in part on a selection model. In accordance with at least one embodiment, the selection model may be based at least in part on a capacity of the storage location and a threshold time duration by which purchased items of the set of items are to be transported from the storage location to a purchaser. A plurality of probability values corresponding to the candidate set of items may be determined. An optimal set of items may be determined based at least in part on the candidate set of items and the plurality of probability values.

Techniques for optimizing an inventory selection

Techniques for managing a removal channel may be described. In an example, a first estimate may be generated. The first estimate may be associated with removing a volume of items from an inventory space within a planning horizon through a removal channel. A decision to remove the volume of items may be accessed. The decision may be based on the first estimate and a second estimate associated with adding a capacity for inventorying an equivalent volume of items. Based on the decision, a constraint may be generated and imposed on an inventory level of an item during a time period of the planning horizon. A quantity of the item to remove within the time period through the removal channel may be estimated based on the constraint. Usage of the removal channel during the time period may be set based on the quantity of the item to remove.

Managing a removal channel based on inventory capacity

Growing community of inventors

Seattle, WA, United States of America

Ali Sadighian