Amazon Technologies, Inc.

Samsara Inc.

Caglar Iskender

Mark Aran Aiken

Jason Frederic Symons

Antonio Vargas Garcia

Karthikram Rajadurai

Arthur Pohsiang Huang

Rhea Lin

Joyce Tian Chen

Milan Krneta

Vignesh Bhuvaneshwar Iyer

Karthikeyan Kailasam

Scott McConnell

Mark Aiken

Ritu Jain

Kushagra Gupta

Samuel Goldman

Timothy Etters

Chris Frost

Adam DeLisse

Brandt Elster

A video streaming system including a video cache system having a cache database and a video storage device, and including a video server configured to communicate with a gateway system configured to store multiple video streams, communicate with a cloud backup system configured to store video in a file system, and communicate with a cloud server to receive a stream request, the stream request indicating requested video that comprises multiple video segments. The video server is further configured to receive the stream request and determine a playlist that indicates where on the video cache system, the cloud backup system, and the gateway system the video segments that are needed to fulfill the stream request are stored, and to communicate needed video segments to the video cache system in preparation for providing streaming video to fulfill the stream request.

Cloud storage backend

Systems and methods for caching on-device commands utilizing utterance patterns are disclosed. For example, user utterances and resulting selected commands are analyzed to determine which utterances result in consistently-selected commands. Pattern data is generated indicating an association between these utterances and commands. The pattern data is stored in an on-device storage and utilized to select a command to be performed when the associated utterances are identified from audio data. When new features are developed for cloud-side services, a determination is made as to whether commands and utterances associated with the new features are candidates for utterance-to-command patterns, and corresponding pattern data may be sent to the on-device storages.

On-device commands utilizing utterance patterns

Techniques for enabling entities to provide a system with data representing specific relationships of various endpoint devices for the purpose of responding to user inputs are described. A system may include an application program interface (API) that enables an entity to provide the system with data describing related endpoint devices, as well as data representing terminology that a user may use to refer to a specific endpoint device. After the system receives such data, when a user associates an endpoint device with their user profile, the system can associate that endpoint device with various other endpoint devices based on data previously received from an entity. When the system receives a user input, the system may traverse endpoint device relationships, represented in a user profile, to determine an endpoint device that should be invoked to respond to the user input.

Endpoint device relationships in processing user inputs

Growing community of inventors

Redmond, WA, United States of America

Caglar Iskender