Amazon Technologies, Inc.

Rohit Malshe

Dipal Patel Gupta

Liron David Yedidsion

Jin Ye

Abhilasha Prakash Katariya

Natarajan Gautam

Julian Enrique Pachon

Yi Liu

Alan Baird

Bhargav Kunkulagunta

Dilip Kumar Sivasankar

Jatin Arora

Daniel Esteban Merchan

Vaibhav Desai

Mahdieh Allahviranloo

Chinmoy Mohapatra

Bingqing Ge

A prompt refinement service can be used to enhance output from a generative model. Entity input that indicates a request for a response can be received. A set of prompt refiners can be received. A prompt refinement service can identify a particular prompt refiner among the set of prompt refiners having a highest similarity to the entity input compared with other prompt refiners. The prompt refinement service can aggregate the particular prompt refiner with the entity input to generate an aggregated input. The prompt refinement service can transmit the aggregated input to the generative model to cause the generative model to return an output in response to the entity input.

Prompt refinement service for enhancing generative model output

A method for supply chain consensus planning using batch computing architecture may include instantiating an instance in a network; based on the instantiation of the instance, writing computational codes of local coordinators from scalable object storage in the network to batch computing devices in the network; (i) executing the computational codes, independently from one another, using the batch computing devices; (ii) based on executing the computational codes, writing first public variables, by the local coordinators independently from one another, from the batch computing devices to the scalable object storage; (iii) identifying, by the instance, based on the first public variables, a first quantity for a time period; iteratively repeating steps (i)-(iii) with updated data from the local coordinators until identifying a convergence on a second quantity, and sending, by the instance, an indication of the convergence to the local coordinators.

Enhanced batch computing architecture and techniques for consensus planning for large-scale supply chains

A system can function relative to an item, a tether, and a robotic manipulator. The tether can correspond to a loop or other structure that can be mountable or mounted in an installed state in which the tether is secured with the item to facilitate lifting the item by lifting of the tether. The robotic manipulator can include a robotic end effector engageable with the tether in the installed state. The robotic end effector can be configurable to an engaged state in which the tether is coupled with the robotic end effector. The robotic manipulator in the engaged state can be operable to move the item by lifting of the tether in the installed state.

Tether usage for robotic package delivery or item handling

A method of optimizing a delivery route for a delivery vehicle driver and a delivery assistant includes: identifying a delivery route for a delivery vehicle; determining, for each of the packages to be delivered using the delivery route, a first delivery time for a delivery vehicle driver of the delivery vehicle to deliver the respective package without the delivery assistant, a second delivery time for a delivery assistant of the delivery vehicle to deliver the respective package without the delivery vehicle driver, and a third delivery time for both the delivery assistant and the delivery vehicle driver to deliver the respective package; determining minimum times needed by the delivery vehicle driver and the delivery assistant to deliver the packages using the delivery route; and determining that a minimum of the estimated total delivery times corresponds to a first subset and a second subset of the packages.

Enhanced dynamic last-mile modeling for delivery assistants

Techniques for dividing a geographical area into districts are described herein. Geospatial vector data, barrier geospatial vector data, road infrastructure data, and historical delivery demand data for a geographical area may be obtained. A plurality of clusters from a stratified sampling of data points for the delivery demand data and barrier penalties from a barrier-aware road graph are generated. A first set of polygons for the plurality of clusters may be generated using a concave hull algorithm. A second set of polygons may be generated using a barrier constrained network Voronoi algorithm that uses the barrier-aware road graph and the first set of polygons as seeds. The second set of polygons may be modified using a bounded Voronoi algorithm that uses a raster cost allocation based on barrier penalties. Coordinates for each polygon of the modified second set of polygons are determined that divide the geographical area.

Systems and methods for road network aware geographical unit generation

Techniques for generating geographical units that can be used to generate delivery routes are described herein. Geospatial vector data and barrier geospatial vector data for a geographical area may be obtained. Seed points for one or more portions of the geographical area may be determined based at least in part on historical delivery volume for the geographical area. A plurality of polygons that represent the geographical area may be determined based at least in part on an algorithm that uses the seed points, the geospatial vector data, and the barrier geospatial vector data. Coordinates for a geographical unit of a plurality of geographical units that divide the geographical area may be determined based at least in part on the plurality of polygons and a polygon-to-polygon barrier aware drive time matrix that identifies a calculated cost for traveling from one polygon to another polygon using barriers identified in the barrier geospatial vector data.

Geographical unit generation

Techniques for a package selection feature for selecting subsets of packages and generating instructions to deliver selected packages are described herein. A model may be generated for recursively determining future forecast for potential deliveries associated with a geographic location based at least in part on capacity constraints, delivery vehicle capacity, and historical delivery data for the geographic location. Information that identifies a set of packages for delivery to the geographic location during a first duration may be received. A value for each subset of a plurality of subsets for the set of packages may be determined based on an algorithm that uses the future forecasts and the information. A particular subset may be selected for delivery to the geographic location for a given carrier during a duration based on an algorithm that uses various parametric values for the particular subset, the future forecasts, and the information.

Growing community of inventors

Bellevue, WA, United States of America

Rohit Malshe