Amazon Technologies, Inc.

Microsoft Technology Licensing, LLC

Sococo, Inc.

Alexander Say Go

Samuel Scott Gigliotti

Tejas Patel

Eric Meissner

Chinmay Milind Dani

Benjamin Markwardt

Michael F Schwartz

Joe Pemberton

Archit Jain

Noah Aaron Cedar Davidson

Vladimir Petter

Vijitha Raji

Donald MacGregor

Steven Mack Saunders

Gregorio Maeso

Tara Raj

Alexandre Joseph Francois Allaire

Donald Macgregor

Techniques for selecting a skill component to process a natural language input are described. When a natural language input is received, natural language understanding (NLU) output data representing the natural language input is generated, and skill components (capable of processing the NLU output data) are determined. Thereafter, rules (for preventing the invocation of skill components) are implemented in a tiered manner, resulting in the determination of a subset of the skill components. The subset of skill components is ranked using a machine learning model(s), and the top-ranked skill component is called to process the NLU output data.

Natural language input routing

This disclosure is directed to generating one or more presence-event notifications at one or more user devices associated with a user, and transmitting those presence-event notifications to one or more presence server(s). At the presence server(s) a stream of presence-event notifications may be filtered according to any number of predefined criteria, and then the presence-event notifications that have filtered through may be applied to presence plug-ins to generate presence information associated with user(s) for whom presence-event notifications were received. The presence-event notifications may be provided as a stream of presence information to one or more consumers of presence information and/or stored in a presence datastore. A consumer of presence information, such as a service provider, may request presence information associated with one or more users. In response, the service provider may be provided with the requested presence information based on whether that service provider has permissions to receive the requested presence information.

Contextual presence

Techniques for performing runtime ranking of skill components are described. A skill developer may generate a rule indicating a skill component is to be invoked at runtime when a natural language input corresponds to a specific context. At runtime, a virtual assistant system may implement a machine learned model to generate an initial ranking of skill components. Thereafter, the virtual assistant system may use skill component-specific rules to adjust the initial ranking, and this second ranking is used to determine which skill component to invoke to respond to the natural language input. Overtime, if a rule results in beneficial user experiences, the virtual assistant system may incorporate the rule into the machine learned model.

Monitoring the health of a computer system and suggesting an order of repair when problems within the computer system have been identified. Problem(s) and problem entity(s) within the computer system are identified during monitoring. Relationship(s) of the problem entities with other entities in the computer system are identified. A relationship type for each of the identified relationship(s) is determined. A combination of the identified problem(s), the identified problem entity(s), and the determined relationship type(s) is analyzed to determine an order in which repairs of one or more user-visible entities of the computing system should occur in order to address the identified problem(s). An alert comprising the determined order of the repairs is then presented to a user.

Automated ordering of computer system repair

Aspects extend to methods, systems, and computer program products for intent-based interactions with cluster resources. One or more computer systems are joined in a computer system cluster to provide defined computing functionality (e.g., storage, compute, network, etc.) to an external system. In one aspect, a data collection intent facilitates collection and aggregation of data to form a health report for one or more components of the computer system cluster. In another aspect, a command intent facilitates implementing a command at one or more components of the computer system cluster. Services span machines of the computer system cluster to abstract lower level aspects of data collection and aggregation and command implementation for higher level aspects of data collection and aggregation and command implementation. Services can be integrated into an operating system to relieve users from having to have operating system knowledge.

Intent-based interaction with cluster resources

Ones of the windows associated with a software process are identified in a screen layout on a local display of a sharer network node. On the sharer network node, an image of the identified windows as they are arranged in the screen layout and free of obscuration by any other windows in the screen layout is generated. The image is transmitted from the sharer network node to a viewer network node. Also, commands derived from input device events on sharer and viewer network nodes are received. The received commands are processed into a command sequence, which is passed to a shared process executing on the sharer network node. Windows associated with the shared process are presented on the sharer network node in accordance with the received command sequence. An image of the presented windows is generated and transmitted from the sharer network node to the viewer network node.

Application sharing

Growing community of inventors

Sammamish, WA, United States of America

Alexander Say Go