Microsoft Technology Licensing, LLC

Chenguang Yang

Julie Christa Seto

Annie Y Bai

Bernhard S J Kohlmeier

Julianne Marie Emmons

James R Fullmer

Bernhard Kohlmeier

Pradeep Chilakamarri

Nathaniel George Freier

Lin Ma

Erez Kikin Gil

Victor Poznanski

Matthew Evan Vogel

Krista Bendig

Madeline Schuster Kleiner

Darron Jack Stepanich

Max Wang

Douglas Maxwell Taylor

Deqing Chen

Micah James Myerscough

Xuan Li

Om Krishna

Mahaveer Bhavarlal Kothari

Jamie Yu

Kyle Von Haden

Mark James Encarnación

Qing Dai

Becky M Morley

Faryal Ali Khan

Michael Louis Urciuoli

Susan Elizabeth Sullivan

Venkateshwar Rao Thota

Peng Qin

Bernhard SJ Kohlmeier

Mark James ENCARNACIÓN

Olivier Destrebecq

Harrison Mackenzie Ting

Alexander Marx Gorischek

A data processing system implements obtaining presentation content for a presentation, the presentation content including a set of slides and a transcript. The system further implements obtaining presentation style information for the presentation from a presentation datastore, generating a prompt for a LLM that includes the transcript and the presentation style information to cause the first LLM to generate an augmented transcript by rewriting the textual cues of the first transcript into a conversational style associated with the presentation style identified in the presentation style information; generating audiovisual content by generating speech for a virtual presenter and generating audiovisual content synchronizing an animation of the virtual presenter with the first audio content, and sending the audiovisual content to the computing devices of users who are in an audience of the presentation.

Artificial intelligence driven presenter

A system and method for identifying relevant content from other documents and presenting candidate documents in a document authoring application is described. The system identifies a pattern of events from the document authoring application. The system receives a request to identify candidate documents pertinent to a document present in the document authoring application. A set of candidate documents is identified from a document library based on the pattern of events, a reuse score for each candidate document, a similarity score for each candidate document, and content in the document present in the document authoring application. The set of candidate documents is provided in a user interface element of the document authoring application.

System for predicting document reuse

Author-created digital agents and content corpora for electronic documents are described. A content corpora service can include application programming interfaces such as for adding content to a corpus, attaching a corpus to a document, retrieving corpora associated with an author, and searching content of a corpus. An author-created digital agent can receive a request from a consumer of a document, determine corpora associated with the document, formulate a query based on the request, and search content associated with the corpora associated with the document.

Content corpora for electronic documents

In non-limiting examples of the present disclosure, systems, methods and devices for exposing computing service skills are presented. A task intent associated with a natural language input may be identified. A plurality of add-in matching language models may be applied to the natural language input, wherein each add-in matching language model corresponds to a computing service skill. A score for each of the computing service skills may be generated. Each of the computing service skills may be ranked. A selectable option to execute an add-in associated with a top ranked computing service skill may be caused to be displayed. An indication that a selection of the option has been made may be received. The add-in associated with the top ranked computing service skill may be executed.

Contextual extensible skills framework across surfaces

A computing device receives a trigger to surface commands. A possible set of commands is identified and broken into categories. A category is surfaced for user interaction.

Context-based command surfacing

Author-created digital agents

A document lens service identifies an end-user targeted to consume a productivity document being authored in a user interface to a productivity application. The service then also identifies a document profile specific to the end-user indicative of a model document for the end-user to consume. While the productivity document is being authored, suggestions may be surfaced in the user interface for aligning the productivity document with the model document.

Enhanced document services

A document lens service identifies a document profile specific to an end-user that is indicative of a model document for the end-user. When a productivity document is opened to be consumed by the end-user in a user interface to a productivity application, the service modifies the productivity document based at least on the document profile to align the productivity document with the model document. The productivity document is then rendered as-modified in the user interface.

A shell application feature can include trigger information and logic. During operation, the shell application feature can receive an indicator that a trigger occurred; and, in response to the trigger, initiate its logic. The logic of the shell feature is locally performed to identify whether state, object type, user actions, or a combination thereof, with respect to the application satisfy a success criteria corresponding to behavior that will be a prerequisite for a potential feature that is not yet fully implemented. If the state, object type, user action, or combination thereof satisfies the success criteria, a success notification can be communicated to a feature collection service.

Feature usage prediction using shell application feature telemetry

A computing device receives a voice command to perform an action within a document. An interpretation of the voice command is mapped to a set of commands. Disambiguation is automatically performed by conducting a user experience to receive additional information.

Intent driven command processing

A computing device receives voice command inputs from a user. The device obtains a language processing result based on the voice command input. The result includes an intent and a set of arguments. The device also obtains a variety of different types of contextual information. An action is identified based on the intent, the arguments, and the contextual information, and the device then suggests the action by displaying a user selectable input mechanism that can be actuated by the user to perform the action. The device can automatically perform the action as well.

Context-based actions

Techniques for an electronic file fetch system are described. A method may comprise sending a request for a roaming file list to a file fetch server from a first client device and receiving the roaming file list from the file fetch server by the first client device. The roaming file list includes a unique identifier associated with an application file stored on a second client device. The unique identifier includes a machine identifier identifying the second client device and a local file path portion identifying the file path of the application file stored on the second client device. Based on the machine state of the second client device as indicated in the roaming file list, the application file stored on the second client device may be downloaded to the first client device. Other embodiments are described and claimed.

File fetch from a remote client device

Growing community of inventors

Redmond, WA, United States of America

Chenguang Yang