Multi-domain Chatbot Automated Agent Intent Routing

The patent introduces an automated agent as an intermediary between the user and multiple domain-specific chatbots. The agent receives user messages, determines intent, identifies relevant domain-specific modules, and routes queries appropriately. This mediator eliminates the need for users to manually navigate through various chatbot modules, significantly reducing the time and effort required to find the appropriate specialized chatbot while maintaining access to domain-specific expertise.

The automated agent serves multiple functions: it acts as a universal interface for all domain-specific chatbots, performs intent classification, routes queries to appropriate domains, and provides fallback responses. This multi-functional approach consolidates what would otherwise require separate systems into a single unified platform that handles diverse user queries across multiple domains efficiently.

The automated agent performs preliminary actions by determining the intent of user messages before routing them to domain-specific chatbots. It pre-identifies the relevant domain-specific module based on intent classification, so that when the query is routed, the appropriate specialized chatbot is already prepared to handle it. This preliminary intent determination and module identification significantly reduces the communication duration while ensuring the query reaches the most reliable specialized chatbot for accurate responses.

The system implements feedback mechanisms where domain-specific modules return responses to the automated agent, which then evaluates whether the response aligns with the user's intent. If the response is satisfactory, it is forwarded to the user; if not, the system can try alternative domain-specific modules or provide null responses. This feedback loop ensures response accuracy while minimizing unnecessary communication steps by quickly identifying and correcting mismatches between user intent and chatbot responses.

The patent segments the complex task of handling user queries into distinct modular components: an automated agent for intent determination and routing, multiple domain-specific modules for specialized processing, and a response evaluation mechanism. Each component has a specific function, making the overall system manageable despite its complexity. The segmentation allows the system to investigate multiple domain-specific modules systematically by dividing the workflow into discrete, independent steps that can be executed sequentially.

The automated agent serves as a mediating layer that manages the complexity of interacting with multiple domain-specific modules. It abstracts the complexity by providing a unified interface for users while handling the intricate logic of intent classification, module selection, and response evaluation behind the scenes. This intermediary shields users from system complexity while enabling comprehensive investigation of relevant domain-specific modules to ensure response alignment with user intent.