Chatbot Real-Time Deployment and Fallback Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing chatbot platforms lack effective methods for real-time conversation insights and deployment, leading to abrupt conversation endings due to fallback skills being triggered, which negatively impacts user experience and satisfaction.

Innovation Solution

The chatbot platform provides real-time deployment and monitoring capabilities, allowing designers to analyze conversations that trigger fallback skills, gain insights, and deploy new versions of chatbots without disrupting ongoing conversations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If fallback skill is triggered when no intent or skill is identified, then the chatbot can handle unexpected inputs, but the conversation ends abruptly and user experience deteriorates

Engineering Contradiction:

Improvechatbot's ability to handle unexpected inputsVSAvoiduser experience and satisfaction

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system implements feedback by capturing conversation logs when fallback skills are triggered and providing them to developers for analysis. This feedback loop enables developers to identify patterns in failed conversations, understand user intents that were not recognized, and continuously improve the chatbot's intent recognition capabilities, thereby resolving the contradiction between handling unexpected inputs and maintaining user experience

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The chatbot platform enables developers to self-diagnose and self-repair issues by providing tools to analyze conversation logs and identify fallback triggers. Developers can independently investigate why fallback skills are triggered and implement improvements without requiring external support, allowing the system to self-correct and maintain both adaptability and reliability

Inventive Principle:

Principle #25Self-service

2Device complexity

If chatbot developer receives feedback only via incidents/tickets, then the feedback mechanism is simple to implement, but the developer cannot effectively figure out how the bot is performing in production environment

Engineering Contradiction:

Improvefeedback mechanism simplicityVSAvoidchatbot performance insights

Core Design Contradiction:

Device complexityVSLoss of information

Solution Approach 1:

The system performs preliminary action by proactively capturing and storing conversation logs with metadata (including fallback triggers) before developers need to analyze them. This pre-capturing of data eliminates the need for complex on-demand data collection mechanisms while ensuring comprehensive performance information is available when developers need to investigate chatbot effectiveness

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary component that sits between the chatbot operations and developers. This intermediary automatically captures conversation logs, processes them to identify fallback triggers, and presents them in an analyzable format. This intermediary layer bridges the gap between simple feedback collection and comprehensive performance insights without requiring developers to build complex analysis systems

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If new version of chatbot is deployed to production, then the chatbot can incorporate improvements, but ongoing conversations may be disrupted

Engineering Contradiction:

Improvechatbot improvement deployment speedVSAvoidconversation continuity

Core Design Contradiction:

ProductivityVSStability of the object's composition

Solution Approach 1:

The system implements dynamic version management where the chatbot can transition between versions based on conversation state. New versions can be deployed and activated selectively - applying improvements to new conversations while maintaining stability for ongoing conversations. This dynamic approach allows continuous deployment without forcing abrupt changes on active users, resolving the contradiction between deployment speed and conversation continuity

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system segments the chatbot deployment process by allowing different versions to coexist and be applied to different conversation contexts. This segmentation enables incremental rollout where improvements can be tested and activated in controlled portions of the system, allowing fast deployment of improvements while maintaining stability for existing conversations through version isolation

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS12231378B2Realtime conversation AI insights and deployment

Publication Date: 2025.02.18 SAP SE

AI summary

A first version of a chatbot using natural language processing conducts conversations with a plurality of users. The chatbot provides responses by triggering a plurality of skills including a fallback skill that is triggered when no other skill corresponds to the intent of the user. A second version of the chatbot is deployed while the first version of the chatbot is concurrently engaged in a set of conversation sessions. The second version of the chatbot is configured to trigger other skills besides the fallback skill for at least a portion of the conversations in which the first version of the chatbot triggered the fallback skill. New conversation sessions are mapped to the second version of the chatbot while the set of conversation sessions the first version of the chatbot is engaged in are still mapped to the first version of the chatbot.