Custom Post-Call Workflow Generation From AI-Monitored Conversations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing contact center systems provide standard post-call workflows that may include unnecessary tasks, tasks already performed during the call, or tasks not applicable to the customer's specific situation, leading to confusion and resource wastage.
Innovation Solution
A machine-based artificial intelligent (AI) system generates a customized, dynamic post-call workflow by monitoring customer-agent interactions, capturing decision-making responses, and optimizing the workflow based on collected information to remove unnecessary tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a standard library of post-call workflow tasks is provided to all customers, then workflow coverage and completeness are improved, but workflow relevance and customer confusion increase due to inclusion of unnecessary or already-performed tasks
Solution Approach 1:
The patent segments the standard workflow library into individual tasks and dynamically assembles a customized subset for each customer based on call monitoring data. This segmentation allows the system to maintain a comprehensive standard library while providing only relevant tasks to each customer, resolving the contradiction between completeness and clarity.
Solution Approach 2:
The system performs preliminary monitoring of the customer-agent call during the interaction to identify which tasks have already been completed or are not applicable. This preliminary action enables the post-call workflow to exclude redundant tasks, improving clarity while maintaining completeness of necessary tasks.
2Quantity of substance
If the agent provides comprehensive post-call workflow instructions during the call, then task completeness is improved, but customer understanding and accurate capture of instructions deteriorate
Solution Approach 1:
The system uses AI monitoring to provide feedback on the customer's understanding and the agent's instructions during the call. This real-time feedback mechanism helps identify tasks that may not have been properly communicated or understood, allowing the post-call workflow to be adjusted accordingly and improving instruction accuracy.
Solution Approach 2:
The AI monitoring system acts as an intermediary between the agent and customer, objectively capturing and analyzing the interaction to determine task completion status. This intermediary role ensures accurate assessment of which tasks were actually completed, reducing information loss in the post-call workflow generation.
3Adaptability or versatility
If the agent provides on-the-fly information not part of the standard workflow (contact numbers, links, tips, processes), then customer service quality is improved, but workflow system accuracy deteriorates due to deviation from standard procedures
Solution Approach 1:
The system dynamically adapts the post-call workflow based on the actual call content monitored by AI. When agents provide on-the-fly information, the monitoring system captures these deviations and adjusts the workflow accordingly, maintaining accuracy by reflecting the actual service provided while preserving the flexibility of agent expertise.
Data Source
AI summary
A communication between parties over a network may be performed to complete a specific workflow and thereby complete a task. Portions of the workflow may be performed during the communication and others performed after the communication has ended. However, a standardized workflow may have variations, such as when portions to complete after the call has ended may have been completed during the communication or when an agent provides additional or alternative tasks. By analyzing the conversation, such as with a neural network or other artificial intelligent system, the portion of the second workflow to be completed after the communication has ended may be produced that accurately reflects the tasks to be completed.