Amazon Technologies, Inc.

Prashant J Thakare

Jeremy Boynes

Shashank Shekhar

Pravi Garg

Andrew Stuart Huntwork

Nicholas Adam Cummings

Gurinder Raju

Shiladitya Roy

Deepak Uttam Shah

Ratika Anand

Parathan Sivagnanasundaram

Xu Zhang

Casey Stuart Smith

Prathyusha Nadella

Janita Shah

Vishnu V Sadhana

Jalaj Kumar

Ganesh Prabu Ravi

Faqin Zhong

Rohit Bhattacharjee

Nikhil Gupta

Karthik Parameswaran

Michael Trinh

Venkat Chakravarthy

Narenda Gyanchandani

Chih-Shin Wang

Nhat Vu Doan

Viraj Vijay Sanghvi

Clare Elizabeth Veladanda

Anna Catherine Bell

Monty Eich

Veer Yuganter Singh

Benjamin Elliott Pew

Sabrina Chandrasekaran

Alekya Jonnala

Sol Jee Cho

David Z Chen

Pranav Kumar Singh

Techniques for resuming a skill session are described. A system receives a user input, determines a skill to execute with respect to the user input, opens a skill session, and causes the skill to execute. After the skill executes, the skill indicates the skill session is to be placed in a background state, resulting in the user being able to conduct one or more skill sessions while the skill processes in the background state. Sometime after placing the skill session in the background state, the skill requests the system to resume the skill session. If the system determines the skill session is to be resumed, the system sends data to the skill indicating that the skill session will be resumed. Once the skill session is to be resumed, the system places the skill session in an active state, and calls the skill to resume processing of the skill session.

Skill session backgrounding and resumption

Techniques for performing speech processing using multi-modal widget information are described. A system may receive input data corresponding to a user input. The system may also receive widget context data corresponding to one or more multi-modal widgets active at a device. The system may use the widget context data to perform natural language understanding (NLU) processing with respect to the user input, and for selecting a skill component for responding to the user input. The system may send a widget identifier to the skill component when invoking the skill to respond to the user input.

Speech processing and multi-modal widgets

Techniques for session-based device grouping are described. In an example, a computer system receives first data requesting an output, determines a function to provide content data based at least in part on the first data, determines a session identifier of a session associated with execution of the function, and determines session data associated with the session identifier. The session data includes a first device identifier of a first device, an indication that the first device is outputting the content data, and a second device identifier of a second device. The computer system also causes, based at least in part on the session data including the second device identifier, the second device to output the content data.

Session-based device grouping

Techniques for providing device functionalities using device components are described. A system receives a system-generated directive from a skill system and determines a workflow to execute. The system implements a response orchestrator that operates based on the workflow that includes interception points where cross-cutting functionalities can be invoked as pluggable components. The interception points occur pre-system-generated directive, pre-device-facing directive, post-device-facing directive generation, post-device-facing directive dispatch, and the like. The system supports asynchronous execution, conditional execution, and sequential execution of components. Data determined by the cross functionality components can be used by other components for processing.

Response orchestrator for natural language interface

A speech-processing system may provide access to one or more skills via spoken commands and/or responses in the form of synthesized speech. The system may be capable of keeping one or more skills active in the background while a user interacts (e.g., provides inputs to and/or receives outputs from) with a skill running in the foreground. A background skill may receive some trigger data, and determine to request the system to return the background skill to the foreground to, for example, request a user input regarding an action previously requested by the user. In some cases, the user may invoke a background skill to continue a previous interaction. The system may return the background skill to the foreground. The resumed skill may continue a previous interaction to, for example, to query the user for instructions, provide an update or alert, or continue a previous output.

Task resumption in a natural understanding system

Disclosed are various embodiments for generating network pages for customers that include customer-generated page portions. A request for a network page is obtained from a client. The network page is associated with a network site hosted on behalf of a first party by a second party. A portion of the network page is obtained from a service operated by the party in response to the request. The network page, which includes the portion, is generated in response to the request. The generated network page is sent to the client in response to the request.

Generating network pages using customer-generated network page portions

Disclosed are various embodiments for securing execution of page generation code. The page generation code may be instrumented with code that facilitates monitoring of one or more resources consumed by the page generation code. Various checks relating, for example, to code size and whether exception-handling code is present may be performed. The page generation code may be executed to generate a network page in response to a request obtained from a client. One or more resource consumption limits may be enforced upon the page generation code.

Securing execution of customer-supplied network page generation code

Disclosed are various embodiments relating to generating portions of network pages associated with a network site, where the portions of the network pages are generated by page generation code that uses a subset of aggregated data provided to it. A request for a network page is received from a first party. Data is aggregated from one or more data sources in response to the request. Each data source provides multiple data items using a multipart encoding. Page generation code supplied by the first party is executed to generate at least a portion of the network page in response to receiving a subset of the data which is used by the page generation code. The page generation code is able to access the subset of the data as predefined variables.

Providing aggregated data to page generation code for network page generation

Disclosed are various embodiments relating to virtual file systems for customers having network sites hosted by a hosting provider. A file system with multiple files of a customer is maintained for a hosting provider. Customer access to a virtualized version of the file system is facilitated. The virtualized version of the file system includes a virtual file that is dynamically generated based at least in part on one of the files in response to an access of the virtual file by the customer. The file system excludes the virtual file. A modification to the virtual file by the customer is obtained. The file in the file system is updated based at least in part on the modification to the virtual file.

Virtual file system for hosted network sites

Disclosed are various embodiments for generating network pages for customers that include customer-generated page portions. A request for a network page is obtained from a client. The network page is associated with a network site hosted by a hosting provider on behalf of a customer. A portion of the network page is obtained from a service operated by the customer in response to the request. The network page, which includes the portion, is generated in response to the request. The generated network page is sent to the client in response to the request.

Disclosed are various embodiments that facilitate error handling in a network resource generation environment. A request for a network resource is obtained from a client. The network resource is associated with a network site hosted on behalf of a first party by a second party. Resource generation code supplied by the first party is executed by a framework in response to the request. A customized error network resource is sent to the client in response to determining that an error has occurred in the framework that executes the resource generation code.

Error handling in a network resource generation environment

Disclosed are various embodiments for facilitating access to data in network page generation code supplied by customers of a hosting provider. A request for a network page is obtained from a client. The network page is associated with a network site hosted by a hosting provider on behalf of a customer. Page generation code supplied by the customer is obtained, and this code includes a dynamic data variable. One or more service calls are executed based at least in part on the dynamic data variable to obtain one or more data objects. The page generation code is executed to generate the network page in response to the request, where the page generation code is executed with the data objects in place of the dynamic data variable.

Facilitating access to data in network page generation code

Growing community of inventors

Mercer Island, WA, United States of America

Prashant J Thakare