Virtual Meeting Proxy Attendance With AI Task Summaries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users unable to attend virtual meetings miss participation opportunities, leading to inefficient discussions, resource-intensive note-taking, and miscommunication due to the absence of input on discussed points.
Innovation Solution
A virtual meeting system that allows absent users to participate via a simulated presence through an AI-generated interface, including a region with their data and AI-generated responses, summaries, and follow-up questions, reducing the need for manual note-taking and additional meetings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If absent users completely disconnect from virtual meetings, then network bandwidth and processing resources are conserved, but participation opportunities are lost and miscommunication occurs
Solution Approach 1:
The system creates a synthetic representation (copy) of the absent user's participation by generating an AI agent that simulates their presence, preferences, and decision-making patterns. This copy allows the absent user to indirectly participate in the meeting without requiring their physical presence or consuming full network resources for actual video/audio transmission.
Solution Approach 2:
An AI agent serves as an intermediary between the absent user and the meeting participants. The agent receives instructions from the absent user, processes meeting discussions, and provides synthesized responses or decisions on their behalf, thereby maintaining information flow without requiring direct connection from the absent user.
2Loss of information
If manual note-taking is assigned to present participants, then absent users receive meeting information, but resource usage increases and discussion efficiency decreases
Solution Approach 1:
The absent user's information needs are served automatically through the AI agent, which autonomously processes meeting content, extracts relevant information, and delivers it back to the absent user without requiring manual note-taking by present participants. The system serves itself by using AI to handle information management tasks.
Solution Approach 2:
The manual mechanical process of note-taking by human participants is replaced with an automated AI-based information processing system. The AI agent electronically captures, processes, and distributes meeting information, substituting the manual workflow with an automated digital system that is more efficient and scalable.
3Loss of information
If additional meetings are scheduled to include absent users, then complete information exchange occurs, but time consumption and scheduling complexity increase
Solution Approach 1:
The AI agent is prepared in advance with the absent user's preferences, decision patterns, and authorization levels. During the meeting, the agent can immediately begin representing the absent user without requiring additional setup or separate meetings, as the preliminary configuration enables real-time participation simulation.
Solution Approach 2:
The functionality of multiple separate meetings (one for present users, another to separately brief absent users) is merged into a single meeting where the AI agent simultaneously participates in the discussion and delivers information to the absent user outside the meeting, eliminating the need for additional time-consuming sessions.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A method includes obtaining input of a first user that has been invited to participate in a virtual meeting. The input indicates the first user is requesting attendance of the virtual meeting by proxy. The input provides first data to be discussed during the virtual meeting. The method includes causing a virtual meeting UI presented during the virtual meeting to include a first region corresponding to the first user. The first region includes at least a portion of the first data. The method includes, during a discussion of the first data during the virtual meeting, detecting an indication of a task to be performed and generating a prompt identifying the task as input to a generative AI model. The method includes causing second data to be presented in the virtual meeting UI. The second data may be based on an output of the generative AI model.