Intelligent customer service training method, device, equipment and medium for dynamic skill mapping

Abstract

The present application discloses an intelligent customer service training method, apparatus, equipment and medium for dynamic skill mapping, comprising: extracting business scenario and skill chain association data from customer service personnel's historical work orders to generate a fine-tuning data set; training a skill recognition agent based on the fine-tuning data set to output a dynamic skill gap matrix of the CSR; constructing a three-dimensional CSR portrait that at least includes cognitive state, behavioral characteristics, and learning preferences; generating a priority training task sequence and special training tasks by a path planning agent; generating personalized test cases by a dialogue simulation agent using the priority training task sequence, special training tasks and three-dimensional CSR portrait; assigning the personalized test cases to target customer service personnel, obtaining business indicator data of the target customer service personnel to quantify the business skill value of the target customer service personnel.

Description

Technical Field

[0001] The present application relates to the field of intelligent training applications, and in particular to an intelligent customer service training method, device, equipment and medium with dynamic skill mapping. Background Art

[0002] Currently, customer service training primarily relies on traditional static courses, questionnaire assessments, and manual supervision. While these methods can improve customer service representative (CSR) skills to a certain extent, they suffer from numerous shortcomings in practice. These include delayed skill diagnosis: Traditional questionnaire assessments and manual supervision fail to capture CSRs' skill deficiencies in real-time in real-world business scenarios, resulting in inadequate and targeted training. Rigid training content: Static course libraries struggle to dynamically adapt to CSRs' foundational competencies and job objectives, failing to meet the individual needs of different roles (e.g., pre-sales, after-sales, and complaint handling). Delayed feedback: Manual supervision and questionnaire assessments are time-consuming and struggle to quantify the correlation between training effectiveness and business metrics (e.g., customer satisfaction and first-time resolution rate), making it difficult to evaluate and optimize training outcomes. Lack of personalization: Existing training methods fail to tailor training to CSRs' cognitive state, behavioral characteristics, and learning preferences, resulting in inconsistent training outcomes. Lack of dynamic update mechanisms: Existing training methods lack dynamic update mechanisms, preventing real-time adjustments to training content and strategies based on CSRs' actual performance and business changes, leading to a disconnect between training content and actual work needs. These defects make it difficult for traditional customer service training methods to meet the needs of enterprises for efficient and accurate customer service training. Summary of the Invention

[0003] The purpose of this application is to provide an intelligent customer service training method, device, equipment and medium with dynamic skill mapping, so as to at least solve the problem that the current traditional customer service training methods are difficult to meet the needs of enterprises for efficient and accurate customer service training.

[0004] To solve the above technical problems, this application provides an intelligent customer service training method with dynamic skill mapping, including:

[0005] Extracting data related to business scenarios and skill chains from customer service personnel's historical work orders to generate a fine-tuning dataset. The fine-tuning dataset contains the correspondence between business scenarios and skill chains, as well as the specific manifestation of each skill in the conversation.

[0006] Training a skill recognition agent based on the fine-tuning dataset to output a dynamic skill gap matrix for the CSR, wherein the dynamic skill gap matrix is used to represent the gap between the performance score of each skill of the customer service representative in the real-time conversation and the preset job skill standard;

[0007] Based on the dynamic skills gap matrix, combined with simulated dialogue test results and real work order behavior data, a three-dimensional CSR profile is constructed, including at least cognitive status, behavioral characteristics, and learning preferences.

[0008] Based on the three-dimensional CSR profile and the dynamic skill gap matrix, the path planning agent generates a priority training task sequence and special training tasks;

[0009] Using the prioritized training task sequence, the specialized training tasks, and the three-dimensional CSR profile, the conversation simulation agent generates personalized test cases;

[0010] The personalized test case is assigned to a target customer service representative, and business indicator data of the target customer service representative is obtained to quantify the business skill value of the target customer service representative.

[0011] Optionally, the step of training a skill recognition agent based on the fine-tuning dataset and outputting a dynamic skill gap matrix of a CSR includes:

[0012] Using thought chain annotation technology to disassemble the skill chains in the fine-tuning training set;

[0013] Input the disassembled skill chain into the skill recognition agent and train the skill recognition agent;

[0014] The skill recognition agent is trained to map training objectives to segmented skills and generate a dynamic skill gap matrix for CSRs.

[0015] Optionally, before constructing a three-dimensional CSR profile comprising at least cognitive status, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and actual work order behavior data, the process further includes:

[0016] Configure simulated dialogue tests and analyze the results to assess the CSR's cognitive status;

[0017] Analyze behavioral data from real work orders and identify behavioral characteristics.

[0018] Optionally, the configuration simulation dialogue test includes:

[0019] Based on the skill deficiency weights in the dynamic skill gap matrix, a specific text in the knowledge base is retrieved using RAG technology to generate an initial dialogue template related to the specific text;

[0020] Based on the characteristics of real business scenarios, industry-specific variables are injected into the initial dialogue template to form a simulated dialogue test script that is adapted to the CSR portrait.

[0021] Optionally, after constructing a three-dimensional CSR portrait comprising at least cognitive status, behavioral characteristics, and learning preferences, the process further includes:

[0022] Periodically collecting behavioral data of the three-dimensional CSR in real work orders, and calculating a deviation value from the simulated dialogue test based on the behavioral data;

[0023] When the deviation value exceeds a preset threshold, the portrait correction mechanism is triggered to reallocate the priority weights of cognitive states and behavioral characteristics in the three-dimensional CSR portrait.

[0024] Optionally, the method further includes:

[0025] Performing simulation testing on the test case to obtain simulation results;

[0026] Input the simulation results into the KPI prediction model, and output the predicted value of the improvement in customer satisfaction after training through the KPI prediction model;

[0027] The predicted value is fed back into the target customer service personnel's business skill value to optimize the test case.

[0028] Optionally, inputting the simulation results into a KPI prediction model and outputting a predicted value of the improvement in customer satisfaction after training through the KPI prediction model includes:

[0029] Normalizing the skill deficiency values in the dynamic skill gap matrix, and calculating the impact coefficient of each skill on the KPI based on changes in customer satisfaction in historical training data;

[0030] The influence coefficient and the current skill deficiency value are weighted and summed to output a quantitative prediction value of the improvement in customer satisfaction after the training.

[0031] To solve the above technical problems, the present application also provides an intelligent customer service training device with dynamic skill mapping, comprising:

[0032] A data extraction module is used to extract business scenarios and skill chain association data from customer service personnel's historical work orders to generate a fine-tuning dataset. The fine-tuning dataset contains the correspondence between business scenarios and skill chains, as well as the specific manifestation of each skill in the conversation;

[0033] A skill matrix module is used to train a skill recognition agent based on the fine-tuning dataset and output a dynamic skill gap matrix for the CSR. The dynamic skill gap matrix is used to represent the gap between the performance score of each skill of the customer service representative in the real-time conversation and the preset job skill standard;

[0034] A profile building module is used to build a three-dimensional CSR profile including at least cognitive status, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data;

[0035] A task planning module, configured to generate a priority training task sequence and special training tasks by a path planning agent based on the three-dimensional CSR profile and the dynamic skill gap matrix;

[0036] A use case generation module, configured to generate personalized test cases by a conversation simulation agent using the priority training task sequence, the special training task, and the three-dimensional CSR portrait;

[0037] The training configuration module is used to assign the personalized test case to the target customer service personnel and obtain the business indicator data of the target customer service personnel to quantify the business skill value of the target customer service personnel.

[0038] In order to solve the above technical problems, the present application also provides a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the above-mentioned dynamic skill mapping intelligent customer service training method.

[0039] In order to solve the above technical problems, the present application also provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the above-mentioned dynamic skill mapping intelligent customer service training method.

[0040] The beneficial effects of the embodiments created by the present application are as follows: a fine-tuning data set is generated by extracting business scenarios and skill chain association data from historical work orders of customer service personnel; a skill recognition agent is trained based on the fine-tuning data set to output a dynamic skill gap matrix of the CSR; according to the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data, a three-dimensional CSR portrait is constructed that at least includes cognitive state, behavioral characteristics, and learning preferences; based on the three-dimensional CSR portrait and the dynamic skill gap matrix, a path planning agent generates a priority training task sequence and special training tasks; using the priority training task sequence, special training tasks and three-dimensional CSR portrait, a dialogue simulation agent generates personalized test cases; the personalized test cases are assigned to target customer service personnel, and the business indicator data of the target customer service personnel is obtained to quantify the business skill value of the target customer service personnel; through job portrait and dynamic skill mapping technology, the skill shortcomings of customer service personnel (CSR) in different business scenarios can be accurately identified, solving the pain point of lagging skill diagnosis in traditional training and realizing real-time dynamic adjustment of training content. Secondly, multimodal learner modeling technology constructs a three-dimensional CSR profile that comprehensively covers cognitive states, behavioral characteristics, and learning preferences, making training more targeted and personalized, effectively improving training effectiveness. Furthermore, virtual-reality-integrated training resource generation technology, combined with a business simulation system, can generate highly realistic, personalized test cases, helping CSRs accumulate experience in a simulated environment and enhance their ability to navigate complex business scenarios. Finally, by quantifying and assessing skill values, the system provides real-time feedback on training effectiveness and predicts the extent of improvement in business KPIs. This addresses the issue of delayed feedback on traditional training results and provides companies with an efficient, accurate, and quantifiable solution for customer service training. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0042] Figure 1 This is a basic flow chart of an intelligent customer service training method with dynamic skill mapping according to a specific embodiment of the present application;

[0043] Figure 2 This is a schematic diagram of the basic structure of an intelligent customer service training device with dynamic skill mapping according to a specific embodiment of the present application;

[0044] Figure 3 This is a basic structural block diagram of a computer device according to a specific embodiment of the present application. DETAILED DESCRIPTION

[0045] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and are not to be construed as limiting the present application.

[0046] Those skilled in the art will understand that, unless otherwise stated, the singular forms "a," "an," "said," and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of this application refers to the presence of the stated features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0047] Those skilled in the art will understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and, unless specifically defined as such, will not be interpreted in an idealized or overly formal sense.

[0048] Those skilled in the art will appreciate that the term "terminal" as used herein includes both devices that are wireless signal receivers, i.e., devices that only have wireless signal receivers without transmission capabilities, and devices that have receiving and transmitting hardware capable of performing two-way communication over a two-way communication link. Such devices may include: cellular or other communication devices with single-line displays, multi-line displays, or cellular or other communication devices without multi-line displays; PCS (Personal Communications Service) devices that may combine voice, data processing, fax, and / or data communication capabilities; PDAs (Personal Digital Assistants) that may include a radio frequency receiver, a pager, Internet / Intranet access, a web browser, a notepad, a calendar, and / or a GPS (Global Positioning System) receiver; and conventional laptop and / or palmtop computers or other devices that have and / or include a radio frequency receiver. As used herein, a "terminal" can be portable, transportable, installed in a vehicle (air, sea, and / or land), or adapted and / or configured to operate locally and / or in a distributed manner at any other location on Earth and / or in space. A "terminal" as used herein can also refer to a communication terminal, an Internet access terminal, or a music / video playback terminal, such as a PDA, a mobile internet device (MID), and / or a mobile phone with music / video playback capabilities, as well as devices such as smart televisions and set-top boxes.

[0049] The hardware referred to by names such as "server", "client", and "service node" in this application is essentially an electronic device with capabilities equivalent to those of a personal computer. It is a hardware device that has the necessary components revealed by the von Neumann principle, such as a central processing unit (including an arithmetic unit and a controller), a memory, an input device, and an output device. Computer programs are stored in its memory, and the central processing unit loads the program stored in the external memory into the internal memory for execution, executes the instructions in the program, and interacts with the input and output devices to complete specific functions.

[0050] It should be noted that the concept of "server" referred to in this application can also be extended to server clusters. Based on the network deployment principles understood by those skilled in the art, the servers described should be logically divided. In physical space, these servers can be independent of each other but callable through interfaces, or integrated into a single physical computer or a computer cluster. Those skilled in the art should understand this flexibility and should not use it to constrain the implementation of the network deployment method of this application.

[0051] Unless expressly specified, one or more technical features of the present application can be deployed on a server for implementation and accessed by a client through a remote call to obtain an online service interface provided by the server, or can be directly deployed and run on a client for implementation.

[0052] Unless explicitly specified, the AI models referenced or may be referenced in this application can be deployed on a remote server and remotely called on the client, or can be deployed and directly called on a client with sufficient device capabilities. In some embodiments, when it runs on the client, its corresponding intelligence can be obtained through transfer learning to reduce the requirements for the client's hardware operating resources and avoid excessive occupation of the client's hardware operating resources.

[0053] Unless explicitly specified, the various data involved in this application can be stored remotely on a server or on a local terminal device, as long as they are suitable for being called by the technical solution of this application.

[0054] Those skilled in the art should be aware that although the various methods of this application are described based on the same concept and thus exhibit commonality, unless otherwise specified, these methods can be independently executed. Similarly, the various embodiments disclosed in this application are all based on the same inventive concept. Therefore, concepts with the same expression, as well as concepts that are appropriately transformed for convenience despite different expression, should be understood as equivalent.

[0055] Unless expressly stated to be mutually exclusive, the various embodiments disclosed in this application may be cross-combined with the relevant technical features of the various embodiments to flexibly construct new embodiments, as long as such combination does not deviate from the creative spirit of this application and can meet the needs of the prior art or resolve certain deficiencies in the prior art. Those skilled in the art should be aware of such flexibility.

[0056] See also Figure 1 , Figure 1 This is a basic flow chart of the intelligent customer service training method with dynamic skill mapping in this embodiment.

[0057] like Figure 1 Shown, including:

[0058] S1100: Extract business scenarios and skill chain association data from customer service personnel's historical work orders to generate a fine-tuning dataset;

[0059] This embodiment can be applied to customer service personnel training scenarios in various fields, including finance, e-commerce, insurance, education, healthcare, law, and hotel management. In this embodiment, a customer service training system is configured to customize and implement systematic and professional training plans for customer service personnel. The system first retrieves historical ticket data for customer service personnel from the enterprise database. This historical ticket data includes the complete conversation text, customer problem description, customer service personnel response, and final solution. To extract data related to business scenarios and skill chains, the system uses natural language processing (NLP) technology to analyze the conversation text. Using pre-trained language models (such as BERT or Transformer architecture), the system can identify key business scenarios in the conversation (e.g., "product inquiry," "refund processing," "complaint handling," etc.). Furthermore, the system can also combine job descriptions (JDs) and quality inspection records to extract the skill chains corresponding to each business scenario. For example, in the "refund processing" scenario, the skill chains might include "payment system operation," "customer complaint reconciliation speech," and "platform policy interpretation." Using annotation technologies (such as CoT), the system maps these skill chains to business scenarios to generate a fine-tuning dataset. This dataset not only contains the correspondence between business scenarios and skill chains, but also records in detail the specific manifestations of each skill in the conversation through annotation, providing high-quality training samples for the subsequent training of skill recognition agents.

[0060] It should be noted that the system of this embodiment can also incorporate multi-source data fusion technology. In addition to historical work order data, the system can also extract relevant information from customer service personnel's performance evaluation records, customer feedback, and training records. For example, performance evaluation records may include performance scores of customer service personnel in specific business scenarios. These scores can serve as a supplement to the skill chain-related data, helping the system to more accurately assess the actual skill level of customer service personnel. By fusing multi-source data, the generated fine-tuning dataset will be more comprehensive and accurate, better reflecting the skill requirements and performance of customer service personnel in different business scenarios.

[0061] It should be noted that the "business scenario" in this embodiment refers to the specific customer problems or request types encountered by customer service personnel in their work, such as "product consultation", "order inquiry", "after-sales service", etc. The "skill chain" refers to the series of skill combinations required to complete a specific business scenario. For example, in the "after-sales service" scenario, the skill chain may include "product knowledge", "communication skills", "problem-solving ability", etc. By associating business scenarios with skill chains, the system can provide customer service personnel with more targeted training content to help them improve their skills in specific scenarios.

[0062] S1200: Training a skill recognition agent based on the fine-tuning dataset to output a dynamic skill gap matrix of the CSR;

[0063] After extracting data related to business scenarios and skill chains from customer service agents' historical work orders to generate a fine-tuning dataset, the system trains a skill recognition agent based on this fine-tuning dataset and outputs a dynamic skill gap matrix for the CSR. Specifically, the system uses the fine-tuning dataset generated in step S1100 to train the skill recognition agent. The core of this agent is a deep learning-based natural language processing model, whose goal is to identify the use of specific skills from customer service agents' conversational text. Specifically, the system uses a pre-trained language model based on the Transformer architecture (such as BERT or GPT) and fine-tunes it to understand the correspondence between business scenarios and skill chains in customer service conversations.

[0064] During training, the system uses conversational text from the fine-tuning dataset as input and annotated skill chains as output labels, training the agent through supervised learning. To improve the model's generalization and accuracy, the system employs data augmentation techniques, such as synonym replacement and sentence reorganization, to generate more diverse training samples. Furthermore, the system incorporates a multi-task learning mechanism, enabling the agent to simultaneously learn to identify skill usage frequency, proficiency, and compatibility with business scenarios. After training, the skill recognition agent analyzes real-time conversations with customer service agents and outputs performance scores for each skill. The system compares these scores with pre-defined job skill standards to generate a dynamic skills gap matrix. This matrix not only demonstrates the gap between the agent's current skills and the target skills, but also dynamically adjusts to changing business scenarios, providing a basis for subsequent personalized training.

[0065] It should be noted that the system of this embodiment can further optimize the training process of the skill recognition agent. For example, it can introduce transfer learning technology to transfer existing general language model knowledge to the customer service domain, reducing training time and resource consumption. Furthermore, the system can employ reinforcement learning mechanisms, allowing the agent to continuously trial and error and optimize in a simulated environment, improving its adaptability to complex business scenarios. Furthermore, the system can incorporate an expert knowledge base, incorporating the experience and knowledge of domain experts into the training process, further improving the agent's accuracy.

[0066] It should be noted that the "Skills Identification Agent" in this embodiment refers to a deep learning-based natural language processing model that identifies the specific skills used by customer service agents from customer service conversation text. The "Dynamic Skills Gap Matrix" is a real-time, updated matrix that shows the gap between a customer service agent's current skill level and the target skills for their position. This matrix can be dynamically adjusted based on changing business scenarios, providing precise guidance for personalized training. In this way, the system can provide each customer service agent with a tailored training plan, helping them quickly improve their skills and better adapt to job requirements.

[0067] S1300: Based on the dynamic skill gap matrix, combined with the simulated dialogue test results and actual work order behavior data, construct a three-dimensional CSR profile that includes at least cognitive status, behavioral characteristics, and learning preferences;

[0068] After training the skill recognition agent based on the fine-tuning dataset and outputting a dynamic skill gap matrix for the CSR, the system constructs a three-dimensional CSR profile based on this dynamic skill gap matrix, combined with simulated conversation test results and real-world ticket behavior data. This profile comprises at least cognitive state, behavioral characteristics, and learning preferences. Specifically, the system first uses the dynamic skill gap matrix output by the skill recognition agent to analyze the gap between the CSR's performance in each skill and the target skill, thereby assessing their cognitive state. For example, through simulated conversation tests, the system can measure the accuracy of the CSR's responses to policy terms and the compliance of their conversational language. These data reflect the CSR's understanding and mastery of business knowledge. Simultaneously, the system extracts behavioral characteristic data from real-world ticket instances, such as response speed, transfer rate, and repeat consultation rate. This data is processed by a data analysis module to assess the CSR's actual work performance. For example, response speed can reflect the CSR's efficiency, while transfer rate reveals their ability to handle complex issues. Furthermore, the system identifies the CSR's learning preferences by analyzing their behavior on the training platform, such as course clickstreams and learning time distribution. For example, by analyzing the number of clicks and dwell time on videos, graphic text, or scenario-based simulations, the system can determine a CSR's preferred learning style. To achieve this, the system employs multimodal data fusion technology, integrating textual data, behavioral data, and preference data. Using machine learning algorithms, the system automatically identifies key features in a CSR's profile and dynamically updates the profile based on real-time data. Ultimately, the system generates a three-dimensional CSR profile that not only comprehensively reflects the CSR's current status but also provides an accurate basis for subsequent personalized training.

[0069] It should be noted that the system of this embodiment can further optimize the construction of CSR profiles by incorporating deep learning technologies. For example, neural networks can be used to perform semantic analysis on simulated conversation test results to more accurately assess the CSR's cognitive state. Furthermore, the system can incorporate time series analysis technology to dynamically monitor behavioral data from real work orders and promptly identify changes in CSR behavior patterns. Furthermore, the system can utilize a user feedback mechanism to allow CSRs to participate in the calibration process of the profile, improving its accuracy and practicality.

[0070] It's important to note that "cognitive status" refers to a CSR's understanding and mastery of business knowledge, assessed through simulated conversation tests and a skills gap matrix. "Behavioral characteristics" refer to a CSR's actual work performance, such as response speed and call transfer rate, and are derived through analysis of real-world work order data. "Learning preferences" refer to a CSR's inclination toward different learning methods during the learning process, derived through analysis of their behavioral data on the training platform. By constructing a three-dimensional CSR profile, the system can provide each CSR with a personalized training plan to help them improve their business capabilities.

[0071] S1400: Based on the three-dimensional CSR profile and the dynamic skill gap matrix, the path planning agent generates a priority training task sequence and special training tasks;

[0072] After constructing a three-dimensional CSR profile encompassing at least cognitive state, behavioral characteristics, and learning preferences based on the dynamic skills gap matrix, combined with simulated dialogue test results and real-world work order behavior data, the path planning agent generates a prioritized training task sequence and specialized training tasks based on the three-dimensional CSR profile and the dynamic skills gap matrix. Specifically, the core of the path planning agent in this embodiment is an intelligent decision-making model based on reinforcement learning. Its goal is to plan the optimal training path based on the CSR's current skill level and job objectives. The path planning agent first analyzes the cognitive state, behavioral characteristics, and learning preferences in the CSR profile and, based on the skill gaps identified in the dynamic skills gap matrix, determines training priorities. For example, if a CSR has significant skill gaps in "emotion recognition" in the "customer complaint handling" scenario and their learning preference is video learning, the path planning agent will prioritize relevant video training courses. When generating the training task sequence, the system employs a hierarchical planning strategy. First, short-term specialized training tasks are generated to address key skill gaps, rapidly improving the CSR's capabilities in specific scenarios. Subsequently, a long-term skill improvement sequence is generated based on the CSR's long-term career development path. For example, for new CSRs, the system initially arranges basic training in product knowledge and communication skills, followed by progressively more complex training in customer relationship management and problem-solving. To ensure the effectiveness of training tasks, a path planning agent monitors the CSR's training progress and performance in real time. If a CSR performs poorly on a specific training task, the system automatically adjusts the task difficulty or adds relevant training modules. Furthermore, the system dynamically adjusts the training task sequence based on the CSR's performance on actual work orders, ensuring that the training content closely aligns with actual work requirements.

[0073] It should be noted that the system of this embodiment can further optimize the decision-making process of the path planning agent by introducing a multi-agent collaboration mechanism. For example, multiple agents can be responsible for task planning in different skill areas, working together to generate a more comprehensive training path. Furthermore, the system can incorporate a user feedback mechanism, allowing CSRs to evaluate the difficulty and practicality of training tasks, thereby optimizing subsequent training content.

[0074] It should be pointed out that the "path planning agent" in this embodiment refers to an intelligent decision-making model based on reinforcement learning, whose function is to generate personalized training paths based on the CSR's skill gaps and profile characteristics. The "priority training task sequence" refers to an ordered list of short-term and long-term training tasks generated based on the CSR's skill shortcomings and job requirements. The "special training task" refers to a targeted training module designed to address the CSR's skill shortcomings in specific business scenarios. Through these technical means, the system can provide CSRs with efficient and personalized training programs to help them quickly improve their business capabilities.

[0075] S1500: Using the priority training task sequence, the special training task, and the three-dimensional CSR portrait, the conversation simulation agent generates personalized test cases;

[0076] After the path planning agent generates a prioritized training task sequence and specialized training tasks based on the three-dimensional CSR profile and a dynamic skills gap matrix, the conversation simulation agent then uses this prioritized training task sequence, specialized training tasks, and the three-dimensional CSR profile to generate personalized test cases. The system uses the conversation simulation agent to generate personalized test cases to help customer service representatives (CSRs) improve their skills in a simulated environment. In one embodiment, the conversation simulation agent determines the difficulty and complexity of the test cases based on the cognitive state and behavioral characteristics of the CSR profile. For example, if a CSR's emotion recognition skills in a "customer complaint handling" scenario are weak, the agent will generate customer conversations containing emotional language to simulate real-world complaint scenarios. The agent also adjusts the format of the test cases based on the CSR's learning preferences. For example, if the CSR prefers video learning, the agent can generate test cases containing video conversations. When generating test cases, the agent extracts accurate policy terms and product information from the enterprise knowledge base to ensure the business accuracy of the test cases. Furthermore, the agent integrates with the business simulation system to simulate different customer types, such as those with regional accents or those with complex complaint profiles, to increase the diversity and challenge of the test cases. The test cases generated by the conversation simulation agent include not only the conversation text but also contextual information, such as customer background and past order records, to help CSRs better understand their needs. Once the test cases are generated, the system dynamically adjusts the difficulty based on the CSR's real-time performance, ensuring the test remains challenging and consistent with the CSR's current skill level.

[0077] It should be noted that the system in this embodiment can further enhance the realism and interactivity of test cases by incorporating multimodal interaction technologies. For example, by combining speech recognition and synthesis technologies, CSRs can engage in voice conversations with simulated customers, improving their communication skills. Furthermore, the system can incorporate sentiment analysis technology to assess CSRs' emotional reactions during conversations in real time, helping them better manage their emotions.

[0078] It's important to note that the "dialogue simulation agent" in this embodiment refers to a natural language generation model based on deep learning. Its function is to generate personalized test cases based on the CSR's profile and training tasks. "Personalized test cases" are simulated conversation scenarios generated based on the CSR's skill weaknesses and learning preferences, designed to help CSRs improve specific skills. Through these technical means, the system can provide CSRs with a highly simulated training environment, helping them better handle various customer issues in their actual work.

[0079] S1600: Allocate the personalized test case to a target customer service representative, and obtain business indicator data of the target customer service representative to quantify the business skill value of the target customer service representative.

[0080] After the conversation simulation agent generates personalized test cases using the prioritized training task sequence, specialized training tasks, and the three-dimensional CSR profile, it assigns these personalized test cases to target customer service personnel and obtains their business indicator data to quantify their business skills. The system assigns the personalized test cases generated in step S1500 to the target customer service personnel (CSRs) and quantitatively evaluates the CSRs' business skills through a series of technical means. First, the system pushes the test cases to the CSRs through an automated allocation mechanism and records various data during their processing, including key business indicators such as response time, problem-solving efficiency, and customer satisfaction. Specifically, to quantify the CSRs' business skills, the system utilizes a multi-dimensional evaluation model. This model combines the CSRs' performance in the test cases with their historical data on real work orders to generate a comprehensive score. For example, the system can analyze the CSRs' conversational text using natural language processing to assess their language compliance, problem-solving skills, and emotional management abilities. The system also calculates their efficiency indicators based on the CSRs' speed and accuracy in handling the test cases. In addition, the system incorporates a business KPI prediction model, correlating and analyzing CSRs' performance in test cases with actual business metrics (such as customer satisfaction and first-time resolution rate). Using machine learning algorithms, the system predicts the extent to which CSRs' business performance will improve after training. For example, if a CSR performs well in a test case, the system can predict the probability of an increase in customer satisfaction in actual work. The system also supports a real-time feedback mechanism, allowing CSRs to immediately receive a detailed skills assessment report upon completing a test case. This report not only displays the CSR's overall score but also provides specific improvement suggestions and follow-up training directions. In this way, the system helps CSRs quickly understand their skill level and continuously optimize training outcomes.

[0081] Furthermore, in certain scenarios, the system can further enhance the accuracy and objectivity of skill assessments by incorporating external evaluation mechanisms. For example, domain experts can be invited to manually review the CSR's performance in test cases and, based on these expert opinions, adjust the system's assessment results. Furthermore, the system can incorporate customer feedback mechanisms, allowing simulated customers to evaluate the CSR's service quality, further improving the assessment system.

[0082] It's important to note that in this embodiment, "business indicator data" refers to key data generated by CSRs while processing test cases, such as response time, problem-solving efficiency, and customer satisfaction, which are used to quantitatively assess the CSR's business skills. "Business skill value" refers to the CSR's comprehensive skill score, calculated using a multi-dimensional assessment model, reflecting the CSR's ability level in specific business scenarios. This allows the system to provide CSRs with accurate skill assessments and improvement suggestions, helping them better address customer issues in their work.

[0083] In a specific implementation scenario, suppose an e-commerce company needs to improve its customer service team's ability to handle order inquiries and after-sales issues during the "Double 11" shopping festival. First, the system extracts historical ticket data from customer service representatives (CSRs) from the company's database, including customer inquiry conversations, order processing records, and after-sales feedback. Using natural language processing, the system analyzes this data, extracting the correspondence between business scenarios (such as "order inquiry" and "return and exchange processing") and required skill chains (such as "order system operation" and "customer soothing techniques"), and generates a fine-tuning dataset. Based on the fine-tuning dataset, the system trains a skill recognition agent. The agent analyzes the CSR's performance in historical tickets, identifies their performance in various skills, and compares this performance against the job skill standards to generate a dynamic skill gap matrix. For example, the system identifies a CSR with significant skill gaps in "explaining return and exchange policies." The system combines the dynamic skill gap matrix with simulated conversation test results and real-world ticket behavior data to construct a three-dimensional profile of the CSR. The profile shows that the CSR's cognitive state in the "Return and Exchange Processing" scenario is "Insufficient Policy Understanding," their behavioral characteristics are "Slow Response," and their learning preference is "Video Tutorials." Based on the CSR profile and the skills gap matrix, the path planning agent generates a prioritized training task sequence and specific training tasks. The system assigns the CSR a specific training task, "Interpreting Return and Exchange Policies," and recommends that they first learn about the policy through a video course before pursuing conversational skills training. The conversation simulation agent generates personalized test cases based on the prioritized training task sequence and the CSR profile. The test case simulates a return and exchange scenario during the "Double 11" shopping festival, where an emotional customer faces the challenge of accurately explaining the policy and calming the customer. After the CSR completes the test case, the system uses natural language processing to analyze the conversation transcript, assessing their conversational compliance and problem-solving skills. It also quantifies their business skill value based on business metrics such as response time. The system generates a detailed skills assessment report, showing the CSR's improvement in their "Interpreting Return and Exchange Policies" skill and predicting the potential increase in customer satisfaction in their actual work. Through this complete training process, CSRs were able to handle customer issues more efficiently during the "Double Eleven" promotion, improving customer satisfaction and business efficiency.

[0084] In the above embodiment, a fine-tuning dataset is generated by extracting business scenarios and skill chain association data from the customer service personnel's historical work orders; a skill recognition agent is trained based on the fine-tuning dataset to output a dynamic skill gap matrix for the CSR; based on the dynamic skill gap matrix, combined with simulated dialogue test results and actual work order behavior data, a three-dimensional CSR profile is constructed that includes at least cognitive state, behavioral characteristics, and learning preferences; based on the three-dimensional CSR profile and the dynamic skill gap matrix, a path planning agent generates a priority training task sequence and special training tasks; using the priority training task sequence, special training tasks, and the three-dimensional CSR profile, a dialogue simulation agent generates personalized test cases; the personalized test cases are assigned to target customer service personnel, and the business indicator data of the target customer service personnel is obtained to quantify the business skill value of the target customer service personnel; through job profile and dynamic skill mapping technology, the skill shortcomings of customer service personnel (CSRs) in different business scenarios can be accurately identified, solving the pain point of lagging skill diagnosis in traditional training and realizing real-time dynamic adjustment of training content. Secondly, multimodal learner modeling technology constructs a three-dimensional CSR profile that comprehensively covers cognitive states, behavioral characteristics, and learning preferences, making training more targeted and personalized, effectively improving training effectiveness. Furthermore, virtual-reality-integrated training resource generation technology, combined with a business simulation system, can generate highly realistic, personalized test cases, helping CSRs accumulate experience in a simulated environment and enhance their ability to navigate complex business scenarios. Finally, by quantifying and assessing skill values, the system provides real-time feedback on training effectiveness and predicts the extent of improvement in business KPIs. This addresses the issue of delayed feedback on traditional training results and provides companies with an efficient, accurate, and quantifiable solution for customer service training.

[0085] In some implementations, S1200 trains a skill recognition agent based on the fine-tuning dataset and outputs a dynamic skill gap matrix of the CSR, including:

[0086] S1211. Disassemble the skill chain in the fine-tuning training set using thought chain annotation technology;

[0087] In this embodiment, when training a skill recognition agent based on the fine-tuning dataset and outputting a dynamic skill gap matrix for a CSR, the system first uses thought chain annotation technology to break down the skill chains in the fine-tuning training dataset. The core of this process is to decompose complex business scenarios into a series of specific skill points, enabling the skill recognition agent to more accurately identify and assess the CSR's skill level. First, the system extracts the correspondence between business scenarios and skill chains from the fine-tuning dataset. For example, in the business scenario of "handling customer complaints," the skill chain may include multiple skill points such as "emotion recognition," "problem identification," and "solution provision." Using thought chain annotation technology, the system breaks down and annotates these skill points in a logical order. For example, the annotation might be "emotion recognition → problem identification → solution provision," forming a clear skill decomposition trajectory. During the annotation process, the system uses natural language processing technology to analyze the conversation text and identify the specific manifestation of each skill point in the conversation. For example, in the "emotion recognition" skill point, the annotator will mark how the CSR identifies the customer's emotional state during the conversation (e.g., through tone, keywords, etc.). This annotation method not only helps the system understand the logical order of skill points, but also provides rich training samples for the skill recognition agent.

[0088] It's important to note that to improve annotation accuracy and efficiency, the system incorporates a semi-automated annotation tool. This tool allows annotators to quickly locate key skill points in conversational text and annotate them. Simultaneously, the system automatically learns annotation rules based on the annotated data, assisting annotators with subsequent annotation tasks. This allows the system to rapidly generate high-quality annotated data, providing a solid foundation for training the skill recognition agent.

[0089] It's important to note that the system can also involve domain experts to further enhance the accuracy and practicality of skill chain decomposition. Domain experts can refine and optimize skill chains based on actual business needs. For example, experts can further break down the skill point of "solution provision" into sub-skill points such as "policy interpretation," "operational guidance," and "follow-up," making the skill chain more detailed and specific.

[0090] Furthermore, the system employs a multi-round annotation and verification mechanism to ensure the quality of the annotated data. After the annotators complete the initial annotation, the system randomly samples a portion of the annotated data for cross-validation. If any inconsistencies are found, the system organizes the annotators to discuss and correct them until the annotation results are consistent. In this way, the system generates high-quality annotated data, providing reliable input for the training of the skill recognition agent.

[0091] S1212: Input the disassembled skill chain into the skill recognition agent to train the skill recognition agent;

[0092] After using thought chain annotation technology to disassemble the skill chains in the fine-tuning training set, this embodiment also inputs the disassembled skill chains into a skill recognition agent to train the skill recognition agent. The core of the skill recognition agent is a natural language processing model based on deep learning. Its purpose is to accurately identify the specific skills demonstrated by customer service representatives (CSRs) in conversational text by learning from the annotated skill chain data. In this embodiment, the skill recognition agent uses a pre-trained language model based on the Transformer architecture (such as BERT or GPT) and performs fine-tuning on this basis. The fine-tuning process includes the following key steps:

[0093] a. Data preprocessing: Clean and format the labeled skill chain data to ensure input data quality. For example, convert the conversation text and corresponding skill labels into a format acceptable to the model.

[0094] b. Model training: The preprocessed data is fed into the pretrained model and trained using supervised learning. The model's goal is to predict corresponding skill labels based on the conversation text. For example, given the conversation text "The customer is emotionally agitated, and the CSR uses soothing words to ease the customer's emotions," the model should output skill labels such as "emotion recognition" and "soothing techniques."

[0095] c. Loss function optimization: The cross-entropy loss function is used to measure the difference between the model prediction value and the true label, and the model parameters are optimized through the back-propagation algorithm to improve the accuracy and generalization ability of the model.

[0096] d. Model Validation and Adjustment: Evaluate the model using the validation set and adjust model hyperparameters, such as learning rate and batch size, based on the evaluation results. Also, introduce an early stopping mechanism to prevent overfitting.

[0097] During training, the system also incorporates data augmentation technology, generating more diverse training samples through synonym replacement and sentence reorganization, further improving the robustness of the model. Furthermore, the system employs a multi-task learning mechanism, allowing the skill recognition agent to simultaneously learn the skill's frequency of use, proficiency, and matching with business scenarios, thereby improving the model's overall performance.

[0098] Furthermore, the system can incorporate transfer learning techniques to transfer existing general language model knowledge to the customer service domain, reducing training time and resource consumption. For example, a language model pre-trained on large-scale text data (such as BERT) can be used and fine-tuned for specific customer service tasks. This allows the pre-trained model to fully leverage its general language knowledge while adapting it to the specific needs of customer service scenarios.

[0099] Furthermore, the system can employ reinforcement learning mechanisms, allowing the skill recognition agent to continuously experiment and optimize within a simulated environment. For example, by interacting with a simulated customer service dialogue system, the agent can adjust its skill recognition strategy based on feedback, thereby improving its adaptability in complex business scenarios. Through these technical means, the system can further enhance the performance of the skill recognition agent, making it more accurate and efficient in real-world applications.

[0100] S1213: Train the skill recognition agent to map the training objectives into segmented skills and generate a dynamic skill gap matrix for the CSR.

[0101] The decomposed skill chains are fed into the skill recognition agent, which is then trained to map training objectives into sub-skills and generate a dynamic skill gap matrix for the CSR. Specifically, the system first decomposes the training objectives set by the company (such as "improving complaint handling efficiency" or "increasing customer satisfaction") into a series of sub-skills. For example, "improving complaint handling efficiency" can be broken down into specific skills such as "emotion recognition," "problem location," and "solution provision." By learning from the annotated skill chains in the fine-tuning dataset, the skill recognition agent can identify the specific manifestations of these sub-skills in conversational text. Next, the agent compares and analyzes the CSR's performance in historical tickets and simulated conversation tests against these sub-skills. Using natural language processing, the system extracts key information from the CSR's conversations and matches it to skill labels. For example, if a CSR fails to effectively identify a customer's emotion when handling a complaint, the agent will flag the CSR's skill as deficient in "emotion recognition." Based on this comparative analysis, the system generates a dynamic skill gap matrix for the CSR. The matrix is presented in tabular form, with the horizontal axis representing each skill segment and the vertical axis comparing the CSR's current skill level to the target skill level. For example, if a CSR's current skill level in "Emotion Recognition" is 30% and the target level is 80%, this skill will appear as a larger gap value in the gap matrix. The system also dynamically adjusts the gap matrix based on the CSR's performance in different business scenarios to reflect changes in their skill level. Finally, the system combines the dynamic skill gap matrix with the CSR profile to provide a basis for subsequent personalized training path planning. In this way, the system can provide each CSR with a customized training plan, helping them quickly improve their skill level and better adapt to job requirements.

[0102] To further optimize the results, the system in this embodiment can further optimize the skills gap matrix generation process by introducing machine learning algorithms. For example, cluster analysis techniques can be used to group CSRs according to skill level and job requirements, generating a more targeted skills gap matrix for each group. This can improve the efficiency of training resource utilization while ensuring that training content is closely aligned with CSRs' actual needs.

[0103] In some embodiments, before constructing a three-dimensional CSR profile comprising at least cognitive status, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and actual work order behavior data, the process further includes:

[0104] S1311. Configure a simulated conversation test and analyze the simulated conversation test results to evaluate the CSR's cognitive status.

[0105] In this embodiment, the system assesses the cognitive status of customer service representatives (CSRs) by configuring simulated conversation tests. These simulated conversation tests are virtual conversation environments designed based on actual business scenarios. They aim to assess a CSR's understanding and application of business knowledge. First, the system designs simulated conversation test cases for various business scenarios based on the company's business needs and job skill requirements. These test cases cover common scenarios such as customer inquiries, complaint handling, and order inquiries, and embed key business knowledge and skills. For example, in the "product inquiry" scenario, test cases might include questions about product features, pricing, and usage instructions to assess the CSR's understanding of product knowledge. Next, the system pushes the simulated conversation test to the CSR and records their performance during the test. Using natural language processing technology, the system analyzes the CSR's conversational text and evaluates the accuracy, completeness, and logic of their responses. For example, the system can check whether the CSR accurately answers customer questions about product features, provides clear solutions, and effectively soothes customers. The system also incorporates a scoring mechanism to rate the CSR's performance in each test case. Scoring criteria include accuracy of responses (e.g., correct citation of policy clauses), compliance of dialogue (e.g., use of standard dialogue templates), and efficiency of problem-solving (e.g., ability to quickly identify issues and provide solutions). Through this comprehensive scoring system, the CSR's cognitive state is quantitatively assessed, providing data support for subsequent profiling. Furthermore, the system generates detailed feedback reports based on the results of simulated conversation tests, identifying areas where the CSR excels and areas requiring further improvement. This feedback not only helps CSRs identify their own knowledge gaps but also provides targeted training recommendations to training managers.

[0106] It should be noted that the system of this embodiment can further enhance the realism and interactivity of simulated conversational testing by introducing a multimodal testing environment. For example, by combining speech recognition and synthesis technologies, the CSR can conduct conversations with virtual customers through voice, simulating a real-world customer service call. Furthermore, the system can incorporate sentiment analysis technology to assess the CSR's emotional reactions during conversations in real time, helping them better manage their emotions.

[0107] S1312. Analyze the behavioral data in the real work order and determine the behavioral characteristics.

[0108] After configuring a simulated conversation test and analyzing the results to assess the CSR's cognitive state, the system analyzes behavioral data from real ticket instances to identify behavioral characteristics. Specifically, real ticket data is a record of CSRs' actual interactions with customers, containing rich behavioral information that reflects their performance in real business scenarios. The system extracts real ticket data handled by CSRs from the enterprise database, including key metrics such as conversation text, handling time, customer satisfaction ratings, ticket transfer rates, and repeat consultation rates. This data is cleansed and formatted using a data preprocessing module to ensure accuracy and consistency. Next, the system uses data analytics to perform multi-dimensional analysis on this behavioral data. For example, the system can evaluate CSR efficiency by calculating the average response time for each ticket; analyze ticket transfer rates to understand CSRs' ability to handle complex issues; and assess CSR service quality by calculating customer satisfaction ratings. Furthermore, the system can use text analysis to extract information from conversation text about the CSR's communication style and problem-solving skills, such as whether they can effectively soothe customers and quickly identify and provide solutions. To more comprehensively assess CSR behavioral characteristics, the system incorporates machine learning algorithms such as cluster analysis and association rule mining. Cluster analysis can group CSRs based on their behavioral patterns, for example, assigning CSRs with high efficiency and high satisfaction to one group and those with low efficiency and high call transfer rates to another. Association rule mining can uncover potential relationships between behavioral data, such as the negative correlation between response time and customer satisfaction. Ultimately, the system integrates the analysis results into a CSR behavioral profile, including key indicators such as response speed, call transfer rate, repeat consultation rate, and customer satisfaction. These behavioral characteristics not only reflect CSRs' actual work performance but also provide important evidence for subsequent personalized training planners.

[0109] Furthermore, the system can further enhance the timeliness and accuracy of behavioral profile assessments by incorporating real-time data analysis technology. For example, the system can monitor the CSR's work order processing process in real time, promptly identifying abnormal behavior and issuing early warnings. Furthermore, the system can be integrated with business simulation systems to simulate complex scenarios encountered in real work orders, further verifying the CSR's behavioral characteristics and problem-solving capabilities.

[0110] Furthermore, the system can incorporate a customer feedback mechanism, allowing customers to evaluate the CSR's service quality. This customer feedback data can be combined with behavioral data to further refine the CSR's behavioral profile. For example, customer evaluations of CSRs can reflect their communication skills, problem-solving abilities, and emotional management. This allows the system to more comprehensively assess CSRs' behavioral characteristics and provide more precise support for companies' customer service training.

[0111] In some implementations, configuring a simulated conversation test includes:

[0112] S1321. Based on the skill deficiency weights in the dynamic skill gap matrix, retrieve specific text in the knowledge base using RAG technology to generate an initial dialogue template related to the specific text;

[0113] In this embodiment, the system uses Retrieval-Augmented Generation (RAG) technology, combined with skill deficit weights from a dynamic skill gap matrix, to retrieve specific text from the enterprise knowledge base and generate relevant initial conversation templates. This process aims to provide accurate and practical conversation content for simulated conversation testing, ensuring that the test can provide targeted training for CSRs' actual skill shortcomings. First, the system identifies the CSR's skills where they have significant gaps based on the skill deficit weights in the dynamic skill gap matrix. For example, if the CSR's skill deficit in "Emotion Recognition" in the "Customer Complaint Handling" scenario has a high weight, the system will prioritize this skill in the search. Next, the system uses RAG technology to retrieve specific text related to this skill from the enterprise knowledge base. The knowledge base stores structured and unstructured data such as corporate policies, product information, and FAQs. RAG technology uses the Retriever module to quickly locate text snippets in the knowledge base related to the skill deficit. These snippets are then passed to the Generator module to generate initial conversation templates related to the specific text. For example, the retrieval module might find a detailed description of the "customer complaint handling process" in the knowledge base. The generation module then uses this information to generate an initial conversation template, encompassing common customer complaint questions, standard CSR responses, and emotional soothing techniques. This generated conversation template not only incorporates business knowledge but also embeds key skills training points, such as "how to identify customer emotions" and "how to effectively soothe customer emotions."

[0114] In this embodiment, to ensure the quality of the generated dialogue templates, the system incorporates an expert review mechanism. Domain experts review and optimize the generated initial dialogue templates to ensure they meet actual business scenarios and training requirements. This ensures that the system generates high-quality initial dialogue templates, providing a solid foundation for subsequent simulated dialogue testing.

[0115] Furthermore, the system in this embodiment can further enhance the quality of retrieval and generation by incorporating multi-source knowledge base fusion technology. For example, in addition to the company's internal knowledge base, the system can also integrate external resources such as industry standard knowledge bases and FAQ repositories to enrich the search content sources. This ensures that the generated dialogue templates not only contain company-specific knowledge but also encompass general industry best practices.

[0116] S1322. Inject industry-specific variables into the initial dialogue template based on the characteristics of real business scenarios to form a simulated dialogue test script that is adapted to the CSR portrait.

[0117] Based on the skill deficit weights in the dynamic skill gap matrix, the system uses RAG technology to retrieve specific text from the knowledge base and generate an initial conversation template related to the specific text. Then, based on the characteristics of real business scenarios, industry-specific variables are injected into the initial conversation template to create a simulated conversation test script tailored to the CSR profile. Specifically, the system injects industry-specific variables into the initial conversation template to generate a simulated conversation test script tailored to the CSR profile. This process ensures that the simulated conversation test closely replicates real business scenarios while also being customized to the individual needs of CSRs. First, the system analyzes the characteristics of real business scenarios and extracts key industry-specific variables. These variables may include industry terminology, business processes, customer types, and frequently asked questions. For example, in the e-commerce industry, variables may include "Double 11 Big Sale," "Return and Exchange Policy," and "Logistics Information Inquiry." In the financial industry, variables may include "Wealth Management Product Recommendations," "Loan Application Process," and "Risk Warnings." Next, the system personalizes the initial conversation template based on the cognitive state, behavioral characteristics, and learning preferences of the CSR profile. For example, if a CSR's "emotion recognition" skill is weak in the "customer complaint handling" scenario and his learning preference is video tutorials, the system will inject more prompts and examples of speech on emotion recognition into the dialogue template, and add relevant video links or animation demonstrations to the test script.

[0118] Furthermore, the system incorporates template engine technology to efficiently inject industry-specific variables into the initial conversation template. The template engine dynamically replaces variable placeholders within the template based on pre-set rules and logic. For example, in a conversation template about "customer complaint handling," the system could inject the variable "customer emotional agitation" into the conversation, generating a simulated scenario of an emotionally agitated customer and requiring the CSR to identify and soothe the customer's emotions during the conversation.

[0119] Furthermore, the system dynamically adjusts the difficulty of the dialogue script based on the CSR's skill level and training goals. For example, for beginners, the script might provide more prompts and guidance; while for experienced CSRs, the script might be more complex, incorporating more variables and unexpected situations. In this way, the system generates highly personalized and challenging simulated dialogue test scripts, helping CSRs improve their skills in a simulated environment.

[0120] Furthermore, the system can enhance the realism and interactivity of simulated dialogue test scripts by incorporating multimodal interaction technologies. For example, by combining speech recognition and synthesis technologies, CSRs can engage in voice conversations with virtual customers, simulating real-world customer service scenarios. Furthermore, the system can incorporate sentiment analysis technology to assess CSRs' emotional responses in real time during conversations, helping them better manage their emotions.

[0121] In some embodiments, after constructing a three-dimensional CSR profile comprising at least cognitive status, behavioral characteristics, and learning preferences, the process further includes:

[0122] S1331: Periodically collect the behavior data of the three-dimensional CSR in the real work order, and calculate the deviation value from the simulated dialogue test based on the behavior data;

[0123] In this embodiment, after constructing a three-dimensional CSR profile encompassing at least cognitive state, behavioral characteristics, and learning preferences, the system also periodically collects behavioral data from real work orders using these three dimensions. Based on this behavioral data, it calculates a deviation from the simulated conversation test results. Specifically, the system periodically collects behavioral data from real work orders using these three dimensions and calculates the deviation between this data and the simulated conversation test results to assess the accuracy and timeliness of the CSR profile. This process is a key step in dynamically updating the CSR profile, ensuring that the profile reflects the CSR's latest performance in actual work. First, the system sets a reasonable data collection cycle, such as weekly or monthly, and regularly extracts real work order data handled by the CSR from the enterprise database. This data includes key behavioral indicators such as response time, transfer rate, customer satisfaction score, and problem-solving efficiency. The system also records the CSR's performance data from the most recent simulated conversation test, including test scores and skill mastery. Next, the system uses a data analysis module to calculate the deviation between the real work order behavioral data and the simulated conversation test results. This deviation can be calculated using various statistical methods, such as mean squared error (MSE) or mean absolute error (MAE). Taking response time as an example, the system calculates the difference between the average response time of CSRs in real work orders and the response time in simulated dialogue tests. If the difference is large, it means that the CSR's actual work performance deviates from the simulated test.

[0124] Furthermore, the system incorporates a weighting mechanism for deviation values, assigning different weights based on the impact of different behavioral indicators on CSR work. For example, customer satisfaction may be given a higher weight because it directly impacts the customer experience and corporate reputation. This weighted calculation enables the system to more accurately assess the overall deviation of the CSR profile. Finally, the system compares the calculated deviation value with a preset threshold. If the deviation value exceeds the threshold, it indicates that the CSR profile may need to be updated to reflect the latest CSR performance in actual work. This provides the trigger for subsequent profile correction mechanisms.

[0125] S1332: When the deviation value exceeds a preset threshold, a portrait correction mechanism is triggered to reallocate the priority weights of cognitive states and behavioral characteristics in the three-dimensional CSR portrait.

[0126] After periodically collecting the three-dimensional CSR behavioral data from real work orders and calculating the deviation from the simulated conversation test, a profile correction mechanism is triggered when the deviation exceeds a preset threshold. This mechanism re-assigns the priority weights of the cognitive states and behavioral characteristics within the three-dimensional CSR profile. This mechanism is triggered when the system detects that the deviation between the CSR profile and actual performance exceeds a preset threshold. The core of this mechanism is to reassess and adjust the priority weights of the cognitive states and behavioral characteristics within the CSR profile to ensure that the profile more accurately reflects the CSR's actual capabilities and work performance. First, the system analyzes the specific sources of the deviations to determine which profile dimensions require adjustment. For example, if the deviation is primarily due to response time, this indicates a potential problem with the CSR's work efficiency, and the weighting of the behavioral characteristic of response speed may need to be adjusted. If the deviation is due to customer satisfaction scores, the CSR's cognitive states, such as their compliance with their conversational language and their ability to manage emotions, may need to be reassessed. Next, the system dynamically adjusts the weighting based on the magnitude and direction of the deviations. For example, if a CSR's customer satisfaction score in a real-world ticket is lower than the simulated test result, the system will increase the weights of "emotional management" and "speech compliance" in the cognitive state, while decreasing the weight of "response speed" in the behavioral characteristic (assuming the response speed is good). This adjustment reflects the skills that the CSR needs to improve in real work. To achieve dynamic weighting adjustments, the system uses an adaptive algorithm to automatically optimize the weight distribution based on the CSR's historical performance and real-time data. For example, the system can incorporate reinforcement learning mechanisms to find the optimal weight distribution through trial and error and feedback. The system also records the results of each adjustment for subsequent analysis and optimization. After the weight adjustment, the system regenerates the CSR profile and assigns personalized training tasks to the CSR based on the new profile. For example, if the adjusted behavioral characteristic weights indicate that the CSR needs to improve "problem-solving efficiency," the system will prioritize relevant specialized training tasks to help the CSR better handle complex issues in real work.

[0127] In some embodiments, the method further comprises:

[0128] S1611: Perform simulation test on the test case and obtain simulation results;

[0129] S1612: Inputting the simulation results into a KPI prediction model, and outputting a predicted value of the improvement in customer satisfaction after training through the KPI prediction model;

[0130] S1613: Feedback-associate the predicted value with the business skill value of the target customer service personnel to optimize the test case.

[0131] In this embodiment, the system verifies and evaluates the generated personalized test cases through simulation testing to obtain simulation results. The purpose of simulation testing is to simulate real business scenarios, evaluate CSR performance in test cases, and provide data support for subsequent KPI prediction and test case optimization. First, the system builds a business simulation environment that simulates the interaction between real customers and CSRs. The simulation environment includes virtual customers (played by a conversation simulation agent), CSRs (played by target customer service personnel), and business scenarios (such as order inquiries and returns and exchanges). The virtual customers will engage in conversations with CSRs according to pre-set conversation scripts, simulating the behavior and needs of real customers. During the simulation test, the system records various CSR performance metrics during the conversation, including response time, problem-solving efficiency, customer satisfaction ratings, and emotional management skills. This data is analyzed using natural language processing technology to generate a detailed simulation results report. For example, the system can analyze whether the CSR is able to effectively soothe customers' emotions when handling customer complaints and whether they can quickly identify problems and provide solutions. To ensure the accuracy and reliability of the simulation test, the system incorporates a multi-round dialogue mechanism. During each conversation, the virtual customer dynamically adjusts the content based on the CSR's responses, simulating the likely reactions of a real customer. For example, if the CSR fails to effectively resolve an issue, the virtual customer might express dissatisfaction and request further solutions. This allows the system to comprehensively evaluate the CSR's performance in different scenarios.

[0132] The system then inputs the simulation test results into a KPI prediction model to predict the magnitude of the improvement in customer satisfaction after CSR training. This process aims to provide quantitative evaluation and forward-looking predictions of training effectiveness through data analysis and machine learning. First, the system extracts key indicators from the simulation test, such as the CSR's response time, problem-solving efficiency, customer satisfaction score, and emotional management skills. These indicators are organized into structured data and input into the KPI prediction model. The KPI prediction model is a machine learning-based prediction model typically constructed using regression analysis, time series analysis, or deep learning algorithms (such as neural networks). The model's training data includes historical ticket data, CSR training records, and actual customer satisfaction feedback. Using this data, the model learns the customer satisfaction performance of CSRs at different skill levels and builds a predictive model. For example, the model can learn that a 10% reduction in CSR response time is likely to lead to a 5 percentage point increase in customer satisfaction, or that improving a CSR's emotional management skills will lead to a specific increase in customer satisfaction. After inputting the simulation results, the KPI prediction model outputs a predicted value for the improvement in customer satisfaction after the CSR completes the training based on its internal prediction logic. For example, if simulation results show that a CSR’s ability to manage emotions when handling customer complaints has significantly improved, the model might predict an 8 percentage point increase in customer satisfaction.

[0133] Furthermore, the system correlates the predicted customer satisfaction improvement output by the KPI prediction model with the CSR's business skill score to optimize test cases. This feedback mechanism ensures that test cases better reflect the CSR's actual performance improvement and provides more accurate guidance for subsequent training. First, the system compares and analyzes the predicted customer satisfaction improvement with the CSR's business skill score in the simulation test. Business skill scores are derived by quantifying the CSR's performance in various skills, such as emotional management, problem-solving efficiency, and compliance with sales pitches. The system uses correlation analysis to identify which skill improvements contribute most to the increase in customer satisfaction. For example, if the prediction results indicate that the increase in customer satisfaction is primarily due to the CSR's improved emotional management skills, the system will mark this skill as a key skill. Next, the system optimizes the test cases based on the correlation results. If certain test cases fail to effectively improve the CSR's key skills, or if the CSR's performance in certain skills does not match the predictions, the system will adjust the test case content and difficulty. For example, if a CSR's performance in emotion management skills does not meet expectations, the system may add more dialogue scenarios on emotion recognition and soothing, or adjust the complexity of the dialogue scenarios to better train the CSR's skills.

[0134] In this way, the system can more comprehensively evaluate the effectiveness of test cases, promptly identify deviations between test cases and actual business needs, and provide more accurate support for the company's customer service training.

[0135] In some embodiments, inputting the simulation results into a KPI prediction model, and outputting a predicted value of the improvement in customer satisfaction after training through the KPI prediction model, further includes:

[0136] S1621. Normalize the skill deficiency values in the dynamic skill gap matrix, and calculate the impact coefficient of each skill on the KPI based on changes in customer satisfaction in historical training data.

[0137] S1622. Take a weighted sum of the influence coefficient and the current skill deficiency value, and output a quantitative prediction value of the improvement in customer satisfaction after the training.

[0138] In this embodiment, the process of inputting the simulation results into the KPI prediction model and outputting a predicted value for the improvement in customer satisfaction after training also includes normalizing the skill deficit values in the dynamic skill gap matrix. Combined with customer satisfaction changes in historical training data, the impact coefficient of each skill on the KPI is calculated, thereby calculating the impact coefficient of each skill on the key performance indicator (KPI). This process aims to quantitatively assess the potential impact of each skill on customer satisfaction and provide data support for subsequent predictions. Specifically, the system first normalizes the skill deficit values in the dynamic skill gap matrix. Normalization converts skill deficit values to a uniform range (e.g., 0 to 1) to enable direct comparison of deficit values across different skills. Normalization methods can include min-max normalization or z-score standardization. For example, for the skill "emotion recognition," deficit values may range from 0 to 100. After normalization, all deficit values are converted to the range of 0 to 1. Next, the system calculates the impact coefficient of each skill on the KPI based on customer satisfaction changes in historical training data. Historical training data includes CSR customer satisfaction scores before and after training, skill improvement, and corresponding business scenarios. The system uses a data analysis module to analyze this data to identify which skill improvements contribute most to increased customer satisfaction. For example, the system may find that improvements in the "emotion recognition" skill contribute 0.8 to increased customer satisfaction, while improvements in the "problem-solving efficiency" skill contribute 0.6. To calculate impact coefficients, the system uses correlation analysis or regression analysis. For example, using a linear regression model, the system can establish a relationship between skill improvement and increased customer satisfaction. The model input is the change in skill deficit value (such as the difference before and after training), and the output is the change in customer satisfaction. Through model training, the system can determine the impact coefficient for each skill. These impact coefficients reflect the sensitivity of each skill to customer satisfaction and provide an important basis for subsequent predictions.

[0139] Furthermore, this embodiment weightedly sums the impact coefficient and the current skill deficit value to output a quantitative prediction of the improvement in customer satisfaction after training. Specifically, the system first performs a weighted summation based on the impact coefficient calculated in step S71, combined with the current skill deficit value. Specifically, the impact coefficient of each skill is multiplied by the corresponding skill deficit value, and the results for all skills are summed. For example, assuming the impact coefficient of the "Emotion Recognition" skill is 0.8 and the current deficit value is 0.6 (normalized), the skill's contribution to customer satisfaction improvement is 0.8 × 0.6 = 0.48. Similarly, after calculating the contribution values of all skills, these values are summed to obtain a total contribution value. Next, the system converts the total contribution value into a quantitative prediction of the improvement in customer satisfaction. For example, if the total contribution value is 1.5, the system can map this value to a specific percentage increase in customer satisfaction based on historical data and business experience. Assuming that historical data shows that a total contribution value of 1.5 corresponds to a 10 percentage point increase in customer satisfaction, the system outputs a prediction value of 10%. To ensure the accuracy and reliability of the predictions, the system incorporates a calibration mechanism. The calibration mechanism compares historical forecasts with actual customer satisfaction increases to adjust the impact coefficients and weightings. For example, if historical forecasts are too high or too low, the system adjusts the impact coefficients based on actual data to improve forecast accuracy. Furthermore, the system can incorporate user feedback mechanisms to allow training managers and CSRs to participate in the forecast evaluation process. For example, the system can display forecasts to training managers and solicit their feedback. Training managers can then adjust the forecasts based on actual business needs and experience. In this way, the system can generate forecasts that better align with actual business needs, improving training effectiveness and user experience.

[0140] Through the above implementation methods, the system can provide training managers with more accurate and timely prediction results, helping them to better plan training content and evaluate training results.

[0141] Please refer to the following for details: Figure 2 , Figure 2 This is a schematic diagram of the basic structure of the intelligent customer service training device with dynamic skill mapping in this embodiment.

[0142] like Figure 2As shown, an intelligent customer service training device with dynamic skill mapping includes: a data extraction module 1100 for extracting business scenarios and skill chain association data from customer service personnel's historical work orders to generate a fine-tuning dataset; a skill matrix module 1200 for training a skill recognition agent based on the fine-tuning dataset and outputting a dynamic skill gap matrix for the CSR; a profile construction module 1300 for constructing a three-dimensional CSR profile that includes at least cognitive state, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data; a task planning module 1400 for generating a priority training task sequence and special training tasks by a path planning agent based on the three-dimensional CSR profile and the dynamic skill gap matrix; a use case generation module 1500 for generating personalized test cases by a dialogue simulation agent using the priority training task sequence, special training tasks, and three-dimensional CSR profile; and a training configuration module 1600 for assigning the personalized test cases to target customer service personnel and obtaining business indicator data of the target customer service personnel to quantify the business skill value of the target customer service personnel.

[0143] The above-mentioned intelligent customer service training device with dynamic skill mapping generates a fine-tuning dataset by extracting business scenarios and skill chain association data from the customer service personnel's historical work orders; trains a skill recognition agent based on the fine-tuning dataset and outputs a dynamic skill gap matrix for the CSR; constructs a three-dimensional CSR profile based on the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data, that includes at least cognitive state, behavioral characteristics, and learning preferences; based on the three-dimensional CSR profile and the dynamic skill gap matrix, a path planning agent generates a priority training task sequence and special training tasks; using the priority training task sequence, special training tasks, and three-dimensional CSR profile, a dialogue simulation agent generates personalized test cases; the personalized test cases are assigned to target customer service personnel, and the business indicator data of the target customer service personnel is obtained to quantify the business skill value of the target customer service personnel; through job profile and dynamic skill mapping technology, the skill shortcomings of customer service personnel (CSRs) in different business scenarios can be accurately identified, solving the pain point of lagging skill diagnosis in traditional training and realizing real-time dynamic adjustment of training content. Secondly, multimodal learner modeling technology constructs a three-dimensional CSR profile that comprehensively covers cognitive states, behavioral characteristics, and learning preferences, making training more targeted and personalized, effectively improving training effectiveness. Furthermore, virtual-reality-integrated training resource generation technology, combined with a business simulation system, can generate highly realistic, personalized test cases, helping CSRs accumulate experience in a simulated environment and enhance their ability to navigate complex business scenarios. Finally, by quantifying and assessing skill values, the system provides real-time feedback on training effectiveness and predicts the extent of improvement in business KPIs. This addresses the issue of delayed feedback on traditional training results and provides companies with an efficient, accurate, and quantifiable solution for customer service training.

[0144] To solve the above technical problems, the present application also provides a computer device. Figure 3 , Figure 3 This is a basic structural block diagram of the computer device in this embodiment.

[0145] like Figure 3As shown, a schematic diagram of the internal structure of a computer device. The computer device includes a processor, a non-volatile storage medium, a memory and a network interface connected via a system bus. Among them, the non-volatile storage medium of the computer device stores an operating system, a database and computer-readable instructions, and the database may store a control information sequence. The processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. The memory of the computer device may store computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor may execute an intelligent customer service training method with dynamic skill mapping. The computer-readable instructions are stored in the form of a computer program. The network interface of the computer device is used to connect and communicate with the terminal. Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0146] In this embodiment, the processor is used to execute Figure 2 The memory stores the program code and various data required to execute the specific functions of the data extraction module 1100, skill matrix module 1200, profile construction module 1300, task planning module 1400, use case generation module 1500, and training configuration module 1600. The network interface is used to transmit data between user terminals or servers. In this embodiment, the memory stores the program code and data required to execute all submodules of the intelligent customer service training device with dynamic skill mapping. The server can call the server's program code and data to execute the functions of all submodules.

[0147] The computer device generates a fine-tuning dataset by extracting business scenarios and skill chain association data from historical work orders of customer service personnel; trains a skill recognition agent based on the fine-tuning dataset and outputs a dynamic skill gap matrix of the CSR; constructs a three-dimensional CSR profile that includes at least cognitive state, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data; generates a priority training task sequence and special training tasks based on the three-dimensional CSR profile and the dynamic skill gap matrix by a path planning agent; utilizes the priority training task sequence, special training tasks, and the three-dimensional CSR profile to generate personalized test cases by a dialogue simulation agent; assigns the personalized test cases to target customer service personnel, obtains business indicator data of the target customer service personnel, and quantifies the business skill value of the target customer service personnel; through job profile and dynamic skill mapping technology, it is possible to accurately identify the skill shortcomings of customer service personnel (CSRs) in different business scenarios, solve the pain point of lagging skill diagnosis in traditional training, and realize real-time dynamic adjustment of training content. Secondly, multimodal learner modeling technology constructs a three-dimensional CSR profile that comprehensively covers cognitive states, behavioral characteristics, and learning preferences, making training more targeted and personalized, effectively improving training effectiveness. Furthermore, virtual-reality-integrated training resource generation technology, combined with a business simulation system, can generate highly realistic, personalized test cases, helping CSRs accumulate experience in a simulated environment and enhance their ability to navigate complex business scenarios. Finally, by quantifying and assessing skill values, the system provides real-time feedback on training effectiveness and predicts the extent of improvement in business KPIs. This addresses the issue of delayed feedback on traditional training results and provides companies with an efficient, accurate, and quantifiable solution for customer service training.

[0148] The present application also provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the intelligent customer service training method with dynamic skill mapping described in any of the above embodiments.

[0149] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed, the program can include the processes in the above-described method embodiments. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0150] Those skilled in the art will appreciate that the steps, measures, and schemes in the various operations, methods, and processes discussed in this application may be interchanged, modified, combined, or deleted. Furthermore, other steps, measures, and schemes in the various operations, methods, and processes discussed in this application may also be interchanged, modified, rearranged, decomposed, combined, or deleted. Furthermore, steps, measures, and schemes in the prior art that are similar to those disclosed in this application may also be interchanged, modified, rearranged, decomposed, combined, or deleted.

[0151] The above description is only part of the implementation methods of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. An intelligent customer service training method with dynamic skill mapping, characterized in that: include: Extracting data related to business scenarios and skill chains from customer service personnel's historical work orders to generate a fine-tuning dataset. The fine-tuning dataset contains the correspondence between business scenarios and skill chains, as well as the specific manifestation of each skill in the conversation. Training a skill recognition agent based on the fine-tuning dataset to output a dynamic skill gap matrix for the CSR, wherein the dynamic skill gap matrix is used to represent the gap between the performance score of each skill of the customer service representative in the real-time conversation and the preset job skill standard; Based on the dynamic skills gap matrix, combined with simulated dialogue test results and real work order behavior data, a three-dimensional CSR profile is constructed, including at least cognitive status, behavioral characteristics, and learning preferences. Based on the three-dimensional CSR profile and the dynamic skill gap matrix, the path planning agent generates a priority training task sequence and special training tasks; Using the prioritized training task sequence, the specialized training tasks, and the three-dimensional CSR profile, the conversation simulation agent generates personalized test cases; The personalized test case is assigned to a target customer service representative, and business indicator data of the target customer service representative is obtained to quantify the business skill value of the target customer service representative.

2. The intelligent customer service training method with dynamic skill mapping according to claim 1, characterized in that: The step of training a skill recognition agent based on the fine-tuning dataset and outputting a dynamic skill gap matrix of a CSR includes: Using thought chain annotation technology to disassemble the skill chains in the fine-tuning training set; Input the disassembled skill chain into the skill recognition agent and train the skill recognition agent; The skill recognition agent is trained to map training objectives to segmented skills and generate a dynamic skill gap matrix for CSRs.

3. The intelligent customer service training method with dynamic skill mapping according to claim 1 is characterized in that: Before constructing a three-dimensional CSR profile comprising at least cognitive status, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data, the following steps are also included: Configure simulated dialogue tests and analyze the results to assess the CSR's cognitive status; Analyze behavioral data from real work orders and identify behavioral characteristics.

4. The intelligent customer service training method with dynamic skill mapping according to claim 3 is characterized in that: The configuration simulation dialogue test includes: Based on the skill deficiency weights in the dynamic skill gap matrix, a specific text in the knowledge base is retrieved using RAG technology to generate an initial dialogue template related to the specific text; Based on the characteristics of real business scenarios, industry-specific variables are injected into the initial dialogue template to form a simulated dialogue test script that is adapted to the CSR portrait.

5. The intelligent customer service training method with dynamic skill mapping according to claim 4 is characterized in that: After constructing a three-dimensional CSR profile that includes at least cognitive status, behavioral characteristics, and learning preferences, the following also applies: Periodically collecting behavioral data of the three-dimensional CSR in real work orders, and calculating a deviation value from the simulated dialogue test based on the behavioral data; When the deviation value exceeds a preset threshold, the portrait correction mechanism is triggered to reallocate the priority weights of cognitive states and behavioral characteristics in the three-dimensional CSR portrait.

6. The intelligent customer service training method with dynamic skill mapping according to claim 1, characterized in that: The method further comprises: Performing simulation testing on the test case to obtain simulation results; Input the simulation results into the KPI prediction model, and output the predicted value of the improvement in customer satisfaction after training through the KPI prediction model; The predicted value is fed back into the target customer service personnel's business skill value to optimize the test case.

7. The intelligent customer service training method with dynamic skill mapping according to claim 6, characterized in that: Inputting the simulation results into the KPI prediction model and outputting the predicted value of the improvement in customer satisfaction after training through the KPI prediction model includes: Normalizing the skill deficiency values in the dynamic skill gap matrix, and calculating the impact coefficient of each skill on the KPI based on changes in customer satisfaction in historical training data; The influence coefficient and the current skill deficiency value are weighted and summed to output a quantitative prediction value of the improvement in customer satisfaction after the training.

8. An intelligent customer service training device with dynamic skill mapping, characterized in that: include: A data extraction module is used to extract business scenarios and skill chain association data from customer service personnel's historical work orders to generate a fine-tuning dataset. The fine-tuning dataset contains the correspondence between business scenarios and skill chains, as well as the specific manifestation of each skill in the conversation; A skill matrix module is used to train a skill recognition agent based on the fine-tuning dataset and output a dynamic skill gap matrix for the CSR. The dynamic skill gap matrix is used to represent the gap between the performance score of each skill of the customer service representative in the real-time conversation and the preset job skill standard; A profile building module is used to build a three-dimensional CSR profile including at least cognitive status, behavioral characteristics, and learning preferences based on the dynamic skill gap matrix, combined with simulated dialogue test results and real work order behavior data; A task planning module, configured to generate a priority training task sequence and special training tasks by a path planning agent based on the three-dimensional CSR profile and the dynamic skill gap matrix; A use case generation module, configured to generate personalized test cases by a conversation simulation agent using the priority training task sequence, the special training task, and the three-dimensional CSR portrait; The training configuration module is used to assign the personalized test case to the target customer service personnel and obtain the business indicator data of the target customer service personnel to quantify the business skill value of the target customer service personnel.

9. A computer device, characterized in that: It includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the intelligent customer service training method with dynamic skill mapping as claimed in any one of claims 1 to 7.

10. A storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the intelligent customer service training method with dynamic skill mapping as claimed in any one of claims 1 to 7.

Images