Proactive Data Caching for Customer Service Call Flows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Customer service interactions often result in dissatisfaction due to the time-consuming process of data retrieval across disparate systems, requiring multiple calls and waiting for responses, which delays task completion.
Innovation Solution
A method that proactively caches data associated with a second state in a customer service interaction call flow when a first state is reached, using machine learning to anticipate data needs and prioritize critical information, making asynchronous calls to pre-fetch data before it is necessary, thereby reducing wait times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If data retrieval is performed on-demand during customer service interactions, then system resource usage is optimized, but customer wait time increases and satisfaction decreases
Solution Approach 1:
The system performs preliminary data retrieval and caching actions before they are strictly needed. When entering a call flow state, the system proactively retrieves data for that state and caches it in advance, so that when the state is actually executed, the data is already available, eliminating customer wait time.
Solution Approach 2:
The system dynamically adjusts data retrieval behavior based on the current call flow state. It calculates a likelihood score for transitioning to future states and adaptively decides whether to pre-fetch data, balancing between proactive caching and resource optimization based on real-time interaction context.
2Loss of time
If proactive data caching is implemented, then customer wait time is reduced, but system resource consumption increases
Solution Approach 1:
The system applies partial pre-fetching based on calculated likelihood thresholds. Instead of caching all possible data, it selectively caches data only when the likelihood of needing it exceeds a threshold, performing just enough preliminary action to reduce wait times without excessive resource consumption.
Solution Approach 2:
The system changes the timing parameter of data retrieval from synchronous (when needed) to asynchronous (in advance), and adjusts the likelihood threshold parameter to control the degree of pre-fetching, optimizing the balance between customer wait time and system resource usage.
3Reliability
If multiple system calls are made for data retrieval, then data accuracy is ensured, but interaction complexity increases and satisfaction decreases
Solution Approach 1:
The system introduces a cache layer as an intermediary between the data sources and the call flow execution. This cache stores pre-retrieved data and serves it quickly when needed, reducing the number of direct system calls required while maintaining data accuracy, thus simplifying the interaction.
Data Source
AI summary
Methods, non-transitory computer readable media and devices are disclosed for caching data associated with a second state in a customer service interaction call flow when a first state in the customer service interaction call flow is reached. For example, a method includes a processor for initiating a customer service transaction associated with a customer service interaction flow in response to a communication from a user, determining that a first state in the customer service interaction flow is reached during the customer service transaction, calculating that a likelihood of needing data associated with a second state in the customer service interaction flow during the customer service transaction exceeds a threshold, retrieving, from a data store, data associated with the second state, and storing data associated with the second state in a cache.