Real-Time Meeting Insight Extraction with Large Language Models
Solution Overview
Problem
Existing collaborative platforms limit the functionality of digital collaboration assistants to capturing transcripts during meetings and generating summaries post-meeting, often including irrelevant information.
Innovation Solution
A digital collaboration assistant that continuously monitors and analyzes meeting contents in real-time, extracting key points, action items, and answering questions, while providing a structured summary log and customized summaries to participants.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the collaboration assistant captures and processes all meeting transcripts after the meeting, then comprehensive meeting records are generated, but the summary includes too much information including irrelevant content
Solution Approach 1:
The patent extracts only the important and relevant meeting contents from the complete transcript using NLP technology. The system identifies and extracts key points, action items, decisions, and other significant information while filtering out irrelevant content, thereby providing a condensed yet comprehensive summary that loses minimal information while reducing overall volume.
Solution Approach 2:
The patent applies different processing qualities to different parts of the meeting transcript. Important contents receive detailed extraction and analysis, while less important contents are summarized or omitted. This local differentiation of processing quality allows the system to maintain high information value while reducing total information volume.
2Loss of information
If the collaboration assistant provides detailed meeting summaries, then comprehensive information is available, but post-meeting processing time increases
Solution Approach 1:
The patent performs preliminary extraction and identification of important meeting contents during or immediately after the meeting using NLP technology. By pre-processing the transcript to identify key information while the meeting context is still fresh, the system reduces the time required for post-meeting processing while maintaining comprehensive information capture.
Solution Approach 2:
The patent replaces manual post-meeting transcript processing with automated NLP-based analysis. This substitution of mechanical human processing with computational algorithms dramatically reduces processing time while maintaining or improving information extraction quality, as the automated system can analyze transcripts instantly without human reading time constraints.
3Productivity
If the collaboration assistant monitors and analyzes all meeting contents in real-time, then relevant insights are provided during the meeting, but system complexity increases
Solution Approach 1:
The patent implements a multi-functional NLP processing system that simultaneously performs transcription, important content extraction, sentiment analysis, action item identification, and summary generation. By consolidating these multiple functions into a single integrated system, the patent reduces overall system complexity compared to having separate systems for each function, while maintaining high meeting efficiency through real-time processing.
Solution Approach 2:
The system automatically monitors, analyzes, and extracts important meeting contents without requiring manual intervention or configuration. The NLP-based collaboration assistant self-adjusts its analysis based on the meeting context, automatically identifying important versus irrelevant content without human input, thereby reducing operational complexity while improving productivity.
Data Source
AI summary
In accordance with examples of the present disclosure, a collaborative platform provides a digital collaboration assistant that continuously monitors and analyzes shared meeting contents (e.g., voice, text chat messages, shared links and documents, presentation materials, and the like) by participants during a collaborative meeting in near real-time, periodically updates a structure summary log of the meeting contents that are deemed important during the collaborative meeting, and interacts with the participants throughout the collaborative meeting in near real-time, for example, to answer questions or provide additional information.