Microsoft Technology Licensing, LLC

Microsoft Corporation

Thomas Matthew Laird-McConnell

Steven Wayne Ickman

Lili Cheng

Christopher Clayton McConnell

William L Portnoy

Christopher Lee Mullins

David Raskino

Gregory Scott Melander

Scott Jeff Gellock

John Anthony Taylor

Daniel James Driscoll

James H Lewallen

Carlos Castro

Qi Lu

Adi Diamant

Sonja Sabina Knoll

Steve Ickman

Craig Jensen

Justin Brooks Cranshaw

Karen Master Ben-Dor

Raz Halaly

Dan Driscoll

Andrea Ryuta Orimoto

Gabriel Gilabert

Daniel Dos Santos Marques

Shunfu Mao

Raymond Edward Ozzie

David Mowatt

William Ramsey

Lars Hasso Liden

Puneet Narula

David Dawson

Christopher Michael Whitten

Arun Keller Krishnamoorthy

Jonas Barklund

Eshchar Zychlinski

Niket K Patwardhan

The disclosure herein describes providing responses to natural language queries associated with transcripts at least by searching multiple indexes. A transcript associated with a communication among a plurality of speakers is obtained, wherein sets of artifact sections are identified in the transcript. A set of section indexes is generated from the transcript based on artifact type definitions. A natural language query associated with the transcript is analyzed using a natural language model and query metadata of the analyzed natural language query is obtained. At least one section index of the set of section indexes is selected based on the obtained query metadata and that selected at least one section index is searched. A response to the natural language query is provided including result data from the searched at least one search index, wherein the result data includes a reference to an artifact section referenced by the searched section index(es).

Providing responses to queries of transcripts using multiple indexes

In non-limiting examples of the present disclosure, systems, methods and devices for enabling conversational bot communications are presented. A marketplace comprising an identity of a plurality of skills may be maintained, each of the plurality of skills associated with a manifest and a conversational bot. A request for execution of a specific skill may be received. A plurality of conversational bots may be identified, each of the plurality of conversational bots being associated with a manifest for executing the specific skill. The plurality of conversational bots may be ranked. The specific skill may be caused to be executed by the highest ranked conversational bot. In some examples, a plurality of highest ranked conversational bots may be surfaced and a bot that is selected may execute the specific skill.

Marketplace for conversational bot skills

A computing system for real-time analyzing meeting conversations using artificial intelligence (AI) is configured to establish a meeting communication channel that facilitates a meeting among multiple client systems. Each of the multiple client systems corresponds to a meeting participant. In response to receiving communication from the multiple client systems, the computing system analyzes currently received communication via an AI engine while the meeting remains active. The currently received communication is communication received within a rolling time window ending at a current time. Based on the analysis, the computing system identifies an agenda item that has a highest correlation with the currently received communication and a sentiment of at least one participant. The computing system then determines a level of interest of the agenda item based on the sentiment of the participant and causes at least one client system to display a notification or an alert or make suggestions.

Conversational AI for intelligent meeting service

The present disclosure relates to processing operations configured to generate and present a chat bot diagnostic visualization that provides real-time visual insights into operation of a chat bot for debugging assistance. A chat bot diagnostic tool is programmed to launch a graphical user interface that displays a visual representation of diagnostic data for debugging a chat bot in real-time. To enable conversational specific data to be reviewed, the present disclosure is configured to implement a middleware adapter within in a run-time stack of a chat bot that is configured to trace data traffic between a chat bot and a channel of an application/service. This enables the chat bot diagnostic tool to observe data traffic and filter the data traffic to identify relevant data for debugging a chat bot. A debugging data visualization of a chat bot interaction is emulated through a graphical user interface of the chat bot diagnostic tool.

Chat bot diagnostic visualization

The disclosure herein describes providing responses to natural language queries associated with transcripts at least by searching multiple indexes. A transcript associated with a communication among a plurality of speakers is obtained, wherein sets of artifact sections are identified in the transcript. A set of section indexes is generated from the transcript based on artifact type definitions. A natural language query associated with the transcript is analyzed using a natural language model and query metadata of the analyzed natural language query is obtained. At least one section index of the set of section indexes is selected based on the obtained query metadata and that selected section index is searched. A response to the natural language query is provided including result data from the searched at least one search index, wherein the result data includes a reference to an artifact section referenced by the searched section index(es).

A method for utilizing shared memory spaces for data lifecycle management in a conversational AI system includes receiving one or more storage requests directing conversational data to multiple nodes of a conversational data object and storing the conversational data that is directed to each different one of the multiple nodes within a different one of multiple discrete memory spaces. Each of the multiple discrete memory spaces is governed by a scope policy defining a lifetime for data stored within the memory space. The method further provides for erasing data of each node of the multiple different nodes of the conversational data object according to the lifetime defined by the scope policy governing the memory space storing the node.

Shared memory space management for conversational artificial intelligence system

A computing system for real-time analyzing meeting conversations using artificial intelligence (AI) is configured to establish a meeting communication channel that facilitates a meeting among multiple client systems. Each of the multiple client systems corresponds to a meeting participant. In response to receiving communication from the multiple client systems, the computing system analyzes currently received communication via an AI engine while the meeting remains active. The currently received communication is communication received within a rolling time window ending at a current time. Based on the analysis, the computing system identifies an agenda item that has a highest correlation with the currently received communication or a task that is assigned to at least one of the participants, and causes, in substantially real time, at least one of the plurality of the client systems to display the agenda item or add the task to a task list.

In non-limiting examples of the present disclosure, systems, methods and devices for enabling conversational bot communications are presented. A natural language input may be received by a first conversational bot. A skill needed to fulfill a request in the natural language input may be identified. A determination may be made that the skill is not executable by the first conversational bot. A determination may be made that the skill is executable by a second conversational bot. A request for execution of the skill may be sent from the first conversational bot to the second conversational bot. A request for information needed for the second conversational bot to execute the skill may be received by the first conversational bot from the second conversational bot. The information may be sent from the first conversational bot to the second conversational bot. The skill may be executed.

Conversational manifests for enabling complex bot communications

Systems, methods, and computer-readable media are presented that, in operation, generate a checklist from content in response to a checklist request from a user. More particularly, in response to the checklist request, a content type of the content identified by the checklist request is determined. The content is then analyzed, according to a determined content type, to identify a set of enumerative elements, the set of enumerative elements comprising at least two enumerative elements. Thereafter, for each enumerative element of the set of enumerative elements, a checklist item is generated, the checklist item corresponding to the enumerative element and including subject matter of the content corresponding to the enumerative element. Further, the generated checklist item is added as an item of the generated checklist. After iterating each enumerative element of the set of enumerative elements, the generated checklist is returned in response to the checklist request.

Checklist generation

A computer-executed conversational system utilizing adaptive dialogs to provide dynamic modification of a conversation structure. The adaptive dialogs facilitate an adaptive and dynamic way to model conversations. A conversational rule may be associated with a trigger received at an adaptive dialog. The conversational rule may populate a plan with steps to provide the conversational structure of the conversational system. Modifying triggers may be received that allow conversational rules to amend the plan including adding steps, removing steps, or modifying steps from the plan. The adaptive dialogs and/or components thereof may be provided as constructible elements in a platform for building a bot such as a developer Software Development Kit (SDK) or the like. In turn, sophisticated conversational modeling may be conveniently facilitated to provide features such as a dialog dispatcher, the ability to handle interruptions, and to build a pluggable, extensible conversational system.

Conversational AI using adaptive dialogs

Conversational event modeling for determining how to process an input in a dialog stack of a computer-executed conversational system. Receipt of an input at an active dialog in a dialog stack may result in generation of an event corresponding to the input. The event may be propagated through the dialog stack. Dialogs in the dialog stack may generate a confidence value in relation to processing the input, and selection of the dialog in the dialog stack for use in processing the input may be based at least in part on the confidence values. In turn, the conversational system may facilitate improved handling of ambiguous or unrelated inputs at dialogs by propagating the event associated with such an input through the dialog stack. The event creation and processing functions may be provided as parameters for dialogs in a modular dialog creation system.

Conversational event modeling

The techniques discussed herein allow an agent that provides dialog-based services to converse with a user via any mode. A central service programmed to be able to communicate via any mode collects information from agents necessary to act on the agents' behalf. In order to ensure that an agent has a maximal presence across modes, the central service provides auxiliary services such as language translation, fidelity upgrading and/or downgrading, modal constraint conformation, etc. Moreover, the central service provides session tracking via user profiles stored at the intercom so that if a user changes a mode by which the user is conversing with an agent, the central service can keep the context of the conversation across the new mode. Furthermore, agents can communicate with other agents via the central service, allowing other agents to provide services to agents or to add their services to an ongoing dialog with a user.

Multi-modal conversational intercom

In an example, a method includes receiving a data package, the data package including a contact identification and a message payload; in response to determining that a key for the contact identification does not exist in a contact table: generating a key for the contact identification; and updating the contact table with the generated key to indicate an association between the key and the contact identification; calculating a hashed identification based on the generated key and contact identification; and associating the message payload with the hashed identification in a message table.

Keyed hash contact table

Many scenarios involve the delivery of messages received from various message sources to recipients, based on subscriptions established between various recipients and message sources. The recipients may be distributively assigned to one or more message servers, either operating independently or as part of an interoperating message server set. Respective message servers may generate a subscription filter indicating whether at least one assigned recipient subscribes to respective message sources, and may utilize the subscription filter to determine, upon receiving a message from a message source, whether to discard the message or store the message for delivery. Interoperating message servers may also exchange subscription filters, and upon receiving a message from any message source, a message server may utilize the subscription filters of the other message servers to determine which other message servers are assigned at least one subscribing recipient, and may retransmit the message only to such message servers.

Growing community of inventors

Kirkland, WA, United States of America

Thomas Matthew Laird-McConnell