Ephemeral Transcript Querying for Privacy-Safe Video Conferences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video conferencing systems face challenges in providing AI services without recording or persisting transcripts, due to privacy concerns, regulatory barriers, and resource inefficiencies, limiting real-time query capabilities.
Innovation Solution
A method where a video conference provider generates a transcript during or after the meeting, uses it for AI queries, and deletes it without persistence, leveraging AI services like GPT language models for near-real-time responses, ensuring confidentiality and reducing resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video conference transcripts are recorded and persisted for AI query processing, then AI service accuracy and completeness are improved, but privacy security and resource consumption worsen
Solution Approach 1:
The patent segments the transcript processing into two distinct phases: an ephemeral phase where transcripts are generated and processed in real-time for AI queries during the meeting, and a deletion phase where transcripts are automatically removed after use. This segmentation allows the system to maintain high AI query accuracy while eliminating privacy security risks associated with persistent storage.
Solution Approach 2:
The patent introduces an intermediary ephemeral transcript storage mechanism that acts as a temporary buffer between the audio stream and AI processing services. This intermediary storage allows complete transcript access for accurate AI responses while being automatically deleted after the meeting concludes, thus mediating between the need for data completeness and privacy protection.
2Adaptability or versatility
If full video conference transcripts are stored persistently, then AI query capabilities are improved, but resource consumption increases
Solution Approach 1:
The patent implements dynamic transcript management where the transcript storage state changes from temporary during the meeting to deleted after the meeting. This dynamic approach allows the system to provide full AI query capabilities when needed while automatically reducing resource consumption by eliminating persistent storage requirements.
Solution Approach 2:
The patent applies the discarding principle by automatically deleting transcripts after they have served their purpose for AI query processing. The system recovers computational resources by removing the need for long-term storage infrastructure, while still maintaining full adaptability for AI queries during the active meeting period.
3Object-affected harmful factors
If transcripts are deleted after meetings, then privacy security is improved, but AI query functionality is limited
Solution Approach 1:
The patent performs preliminary AI query processing and transcript analysis during the meeting while the transcript is still available in ephemeral storage. This preliminary action ensures that all AI query functionalities are fully operational during the meeting, and the transcript is only deleted after it has completed its functional purpose, thus maintaining both privacy security and AI query versatility.
Data Source
AI summary
Techniques for video conference transcript querying using artificial intelligence are provided. In an example method, a video conference provider joins a client device to a video conference. The video conference provider receives a deletion election. The video conference provider receives an audio stream from the client device and generates a portion of a transcript of the video conference. Prior to the video conference concluding, the video conference provider processes the portion of the transcript to configure a large language model (“LLM”) to respond to queries based on the video conference. The video conference provider receives a query relating to the video conference and causes the LLM to process the query and the portion of the transcript. The video conference provider outputs a response, generated by the LLM, to the query. The video conference provider then deletes the portion of the transcript based on the deletion election.