Context-Aware Conversation Assistance with Semantic Search and Generative AI
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing chatbots, whether rules-based or AI-driven, struggle to provide efficient and context-aware responses in live-chat scenarios, especially when switching contexts, and human agents need assistance to quickly address inquiries.
Innovation Solution
A method leveraging semantic search and generative AI to enhance conversation tools by analyzing user queries and conversation history, generating enriched queries, and providing context-aware responses, while masking sensitive information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If intent-based bots are used to classify user input and activate corresponding skills, then the bot becomes more robust, but it loses flexibility to context switches once an intent is activated
Solution Approach 1:
The system dynamically switches between different bot types based on conversation context. It can transition from intent-based bot (for structured tasks) to rules-based bot (for context switching) and to human agents (for complex issues), making the overall system both robust and flexible
Solution Approach 2:
The conversational system is designed to perform multiple functions through different bot types and human agent integration. It handles both structured intent-based tasks and unstructured context switching, providing a universal solution that covers various conversation scenarios
2Adaptability or versatility
If human agents handle conversations directly, then context switching flexibility is maintained, but response time and efficiency decrease due to manual composition of answers
Solution Approach 1:
The system performs preliminary actions by generating draft responses using the intent-based bot before human agents need to compose answers. This pre-computation of potential responses significantly reduces the time agents need to spend formulating answers while maintaining flexibility
Solution Approach 2:
The system introduces an intermediary layer that acts as a assistant to human agents. This intermediary generates suggested responses based on conversation context and intent classification, allowing agents to quickly review and modify rather than compose answers from scratch
3Ease of manufacture
If rules-based chatbots are used with pre-designed decision trees, then the system is easy to build, but it becomes brittle and does not scale
Solution Approach 1:
The system merges rules-based chatbots (for simple, structured conversations) with AI-based chatbots (for complex, unstructured conversations). This combination allows the system to maintain ease of building through rules-based components while achieving scalability through AI components that can handle diverse scenarios
Data Source
AI summary
A method of generating content in association with an information search and retrieval system. It begins by receiving a query from a user. The query is semantically-searched to identify a context. A conversation history between the user and the system is identified. An enriched query is then generated by associating to the query both the context and at least a portion of the conversation history. The enriched query is then evaluated/processed by a generative-AI. In response, information associated with the enriched query is received from the generative-AI. A response to the query is then generated using the information, e.g., by passing the information back to the user, by modifying (e.g., editing or supplementing) the information to generate modified information and passing the modified information back to the user, or by dismissing the information. If sensitive information is identified in the utterance, it is masked prior to generating the enriched query.