Suggested Message Replies With Structured Context Prompting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generating new content for messages, such as emails or chat messages, can be time-consuming due to the need to understand the context of previous messages and determine an appropriate response.
Innovation Solution
A message generation system using a generative AI model that incorporates inferred context, including message thread history, sender and recipient profiles, and user communication styles to suggest draft replies, optimizing prompts for relevance and reducing latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a generative AI model is used to generate message replies, then the quality and context-awareness of suggested replies is improved, but the time required to generate replies increases
Solution Approach 1:
The prompt is segmented into multiple distinct components: message thread history, sender profile information, recipient profile information, and user communication style preferences. This segmentation allows the system to organize and process context information efficiently, providing the generative AI model with structured inputs that improve reply quality while managing generation time through systematic information delivery.
2Manufacturing precision
If extensive context information is included in the prompt, then the relevance and personalization of suggested replies is improved, but the complexity of the system increases
Solution Approach 1:
The system employs a universal prompt structure that can accommodate multiple types of context information (thread history, sender profiles, recipient profiles, communication styles) through a standardized format. This multi-functional approach allows the same system architecture to handle diverse information types without increasing complexity, as all data follows a consistent schema that the generative AI model can process uniformly.
Solution Approach 2:
A prompt construction intermediary layer is introduced between the context data sources and the generative AI model. This intermediary systematically organizes raw context information into a structured prompt format, managing the complexity of integrating multiple data sources while maintaining relevance. The intermediary acts as a mediator that translates diverse context information into a unified input structure for the AI model.
3Speed
If prompt optimization is performed to reduce latency, then the speed of reply generation is improved, but the quality of context utilization may deteriorate
Solution Approach 1:
Context information is pre-processed and organized into the prompt structure before being submitted to the generative AI model. Message thread history, profile information, and communication styles are gathered and formatted in advance, allowing the model to focus on generation rather than information assembly. This preliminary action reduces latency by eliminating processing delays during reply generation while maintaining full context utilization quality.
Data Source
AI summary
Systems and methods for using a generative artificial intelligence (AI) model to generate a suggested draft reply to a selected message. A message generation system and method are described that use inferred context to improve the suggested draft reply message for the user. Various message data and additional context are obtained and included in a prompt provided to the AI model to improve suggested content. In some examples, the message data and additional context include a message thread history and previously sent messages, profile information of the sender and recipient(s) of the selected message, known relationship information between the sender and the user, etc. For instance, the user's preferred communication style and talking points can be inferred based on the profile data, relationship data, and the user's past communications with similar participants and used to tailor the suggested draft reply to the user.