Virtual Agent Integration in Multiparty Conversations

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional virtual agents are designed for one-on-one interactions, making them inefficient for group activities where multiple users need to collaborate and share information, requiring users to switch between group conversations and virtual agent interactions, leading to awkward and error-prone experiences, especially on mobile devices where only one application can run at a time.

Innovation Solution

A virtual agent can be directly invoked and integrated into multiparty conversations, allowing multiple users to interact within the context of the conversation, providing recommendations and performing tasks without requiring users to leave the conversation or act as intermediaries, and storing records of interactions for future reference.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If virtual agents are designed for one-on-one interactions, then the interaction model is simple and manageable, but the system becomes inefficient for group activities requiring multiple users to collaborate

Engineering Contradiction:

Improveefficiency for group activitiesVSAvoidinteraction model complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The virtual agent is designed to handle both one-on-one and multiparty interactions through a unified interface. The same virtual agent can serve individual users or entire groups, adapting its behavior based on the conversation context without requiring separate systems for different interaction types.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent segments the interaction model into distinct layers: the communication layer handles individual message exchanges, while the group layer manages multiparty coordination. This segmentation allows the system to process simple one-on-one interactions efficiently while also supporting complex group activities when needed.

Inventive Principle:

Principle #1Segmentation

2Reliability

If users switch between group conversations and virtual agent interactions, then separate functionality is maintained, but the user experience becomes awkward and error-prone

Engineering Contradiction:

Improveuser experience qualityVSAvoidoperational simplicity

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The patent merges group conversation and virtual agent interaction into a single unified interface. Users can invoke the virtual agent directly within the group conversation context, eliminating the need to switch between applications. The virtual agent processes requests while maintaining awareness of the group conversation context, reducing errors and improving user experience.

Inventive Principle:

Principle #5Merging (Combining)

3Adaptability or versatility

If only one application can run at a time on mobile devices, then resource constraints are managed, but seamless virtual agent participation in group interactions is hindered

Engineering Contradiction:

Improvevirtual agent integration capabilityVSAvoidnumber of running applications

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The virtual agent is nested within the group conversation application rather than running as a separate application. The virtual agent functionality is embedded as a component or service within the existing communication app, allowing it to operate without requiring additional application slots on mobile devices while maintaining full virtual agent capabilities.

Inventive Principle:

Principle #7Nested doll (Nesting)

Data Source

PatentUS9262175B2Systems and methods for storing record of virtual agent interaction

Publication Date: 2016.02.16 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Systems, methods, and apparatus for use with at least one virtual agent. In some embodiments, at least one processor is programmed to store a receipt for an interaction between the at least one virtual agent and one or more users, wherein the receipt comprises at least some information provided by the one or more users to the at least one virtual agent during the interaction.