ML Action Proposal for Customer Service Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Call center representatives face challenges in consistently and effectively addressing customer issues due to varying problem solutions and the difficulty in handling rare or complex issues, especially for inexperienced representatives.
Innovation Solution
A system utilizing trained machine learning models to embed natural language issue statements into vectors, calculate inner products with actions matrices, and generate lists of proposed actions based on probability thresholds, aiding customer service representatives in providing reliable and consistent solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If machine learning models are used to embed natural language issue statements and generate proposed actions, then the reliability and consistency of customer service responses is improved, but the device complexity increases
Solution Approach 1:
The patent introduces machine learning models as intermediary components between the customer service representative and the customer issue. The issue MLM embeds natural language statements into vectors, and the action MLM generates proposed actions based on these embeddings. This intermediary processing layer standardizes and consistentizes the response generation process, improving reliability while managing complexity through modular model architecture.
2Measurement precision
If machine learning models process natural language statements through embedding and inner product calculations, then the precision of action recommendations is improved, but the loss of time increases
Solution Approach 1:
The system performs preliminary actions by pre-training the machine learning models on extensive customer service data before actual use. The issue MLM is pre-trained to effectively embed various natural language issue statements into meaningful vectors, and the action MLM is pre-trained to generate accurate proposed actions. This preliminary training phase enables faster and more precise real-time processing during actual customer service interactions.
Data Source
AI summary
A method including embedding, by a trained issue MLM (machine learning model), a new natural language issue statement into an issue vector. An inner product of the issue vector with an actions matrix is calculated. The actions matrix includes centroid-vectors calculated using a clustering method from a second output of a trained action MLM which embedded prior actions expressed in natural language action statements taken as a result of prior natural issue statements. Calculating the inner product results in probabilities associated with the prior actions. Each of the probabilities represents a corresponding estimate that a corresponding prior action is relevant to the issue vector. A list of proposed actions relevant to the issue vector is generated by comparing the probabilities to a threshold value and selecting a subset of the prior actions with corresponding probabilities above the threshold. The list of proposed actions is transmitted to a user device.