Seat personnel processing method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, can be applied to service system platforms of medical health, financial science and technology and the like, and discloses an agent personnel processing method and device, computer equipment and a storage medium. Carrying out cleaning and normalization preprocessing on the obtained historical job performance data, and carrying out feature extraction on the preprocessed historical job performance data to obtain a feature vector; based on the feature vector, generating a call volume prediction result of the seat queue through a pre-trained call volume prediction model; determining a task distribution strategy according to the call volume prediction result and the current working state of each seat person in the seat queue, and automatically distributing a task to each seat person according to the task distribution strategy; therefore, task distribution and personnel management of the seat personnel can be optimized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a method and device for processing a seat staff, a computer device and a computer readable storage medium. BACKGROUND

[0002] At present, with the development of modern communication technology, the call center has become one of the important channels for enterprises to communicate with customers. The efficient operation of the call center is of great significance to improve customer satisfaction, optimize the use of enterprise resources and enhance market competitiveness. The seat staff is the core resource of the call center, and their work efficiency and state directly affect the service quality and operating cost of the call center. The traditional seat staff management method is mainly based on manual scheduling and simple rule engine, and lacks comprehensive analysis of the historical work performance and real-time work state of the seat staff. At present, the existing management method has the problem of unreasonable task allocation, that is, it is impossible to accurately allocate tasks according to the actual ability and current state of the seat staff, resulting in some seat staff working overload, while some seat staff are in idle state.

[0003] In the field of medical health, the call center undertakes important functions such as patient consultation, appointment registration, medical follow-up, etc. With the increasing attention to health and the increasing shortage of medical resources, the business volume of medical call centers is showing a rapid growth trend. However, the traditional seat staff management method cannot accurately allocate tasks according to the actual ability and current state of the seat staff, which may delay the best opportunity for patients to seek medical treatment and affect the timeliness and effectiveness of medical services.

[0004] In the field of financial technology, the call center is an important window for financial institutions to communicate and serve customers, involving customer consultation, business handling, risk warning, etc. With the rapid development of financial technology, customers' demand for financial services is increasingly diversified and personalized, and higher requirements are put forward for the service quality and response speed of the call center. However, the traditional seat staff management method also faces the problem of being unable to accurately allocate tasks according to the business ability and real-time state of the seat staff in the application of financial technology call center, resulting in low efficiency of part of the business processing and long waiting time of customers.

[0005] Therefore, how to provide a seat staff processing method, device, computer device and computer readable storage medium can optimize the task allocation and personnel management of seat staff, which is a problem that the technical personnel in the field are eager to solve at present. SUMMARY

[0006] In view of the deficiencies of the prior art described above, the purpose of the present application is to provide a method and device for processing agents, computer equipment and computer readable storage medium, aiming to solve the problem of how to optimize the task allocation and personnel management of agents.

[0007] In order to achieve the above purpose, the present application adopts the following technical solutions:

[0008] In a first aspect, the present application provides a method for processing agents, comprising:

[0009] obtaining historical job performance data of all agents in an agent queue;

[0010] performing preprocessing of cleaning and normalization on the obtained historical job performance data, and performing feature extraction on the preprocessed historical job performance data to obtain a feature vector;

[0011] generating a call volume prediction result of the agent queue based on the feature vector through a pre-trained call volume prediction model;

[0012] determining a task allocation strategy according to the call volume prediction result and the current working state of each agent in the agent queue, and automatically assigning tasks to each agent according to the task allocation strategy.

[0013] In a second aspect, the present application provides a device for processing agents, comprising:

[0014] a data acquisition module configured to obtain historical job performance data of all agents in an agent queue;

[0015] a feature extraction module configured to perform preprocessing of cleaning and normalization on the obtained historical job performance data, and perform feature extraction on the preprocessed historical job performance data to obtain a feature vector;

[0016] a result generation module configured to generate a call volume prediction result of the agent queue based on the feature vector through a pre-trained call volume prediction model;

[0017] an agent management module configured to determine a task allocation strategy according to the call volume prediction result and the current working state of each agent in the agent queue, and automatically assign tasks to each agent according to the task allocation strategy.

[0018] In a third aspect, the present application provides a computer equipment comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for processing agents as described above when executing the computer program.

[0019] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the agent processing method.

[0020] Compared with the prior art, the present application provides an agent processing method, device, computer equipment and computer readable storage medium, wherein the historical work performance data of all agents in an agent queue is obtained; the obtained historical work performance data is preprocessed by cleaning and normalization, and the historical work performance data after preprocessing is subjected to feature extraction to obtain a feature vector; based on the feature vector, a call volume prediction model is pre-trained to generate a call volume prediction result of the agent queue; a task allocation strategy is determined according to the call volume prediction result and the current working state of each agent in the agent queue, and tasks are automatically assigned to each agent according to the task allocation strategy; thereby the task allocation and personnel management of agents can be optimized by the present application. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0022] Figure 1 An application environment schematic diagram of an agent processing method provided by an embodiment of the present application.

[0023] Figure 2 A flowchart of an agent processing method provided by an embodiment of the present application.

[0024] Figure 3 A program module schematic diagram of an agent processing device provided by an embodiment of the present application.

[0025] Figure 4 A structure schematic diagram of a computer equipment provided by an embodiment of the present application.

[0026] Figure 5 Another structure schematic diagram of a computer equipment provided by an embodiment of the present application. DETAILED DESCRIPTION

[0027] With reference to the drawings and the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are a part rather than all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts should fall within the scope of the present application.

[0028] It should be understood that the term "comprising" as used in the specification and the appended claims indicates the presence of the described 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.

[0029] It should also be understood that the term "and / or" as used in the specification and the appended claims indicates any combination of one or more of the associated listed items and all possible combinations of the items.

[0030] As used in the specification and the appended claims, the term "if" can be interpreted as meaning "when" or "once" or "in response to a determination" or "in response to detecting" depending on the context. Similarly, the phrase "if determined" or "if detected [the described condition or event]" can be interpreted as meaning "once determined" or "in response to a determination" or "once detected [the described condition or event]" or "in response to detecting [the described condition or event]" depending on the context.

[0031] In addition, in the description of the specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0032] The reference "one embodiment" or "some embodiments" and the like described in the present specification means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Therefore, the statements "in one embodiment", "in some embodiments", "in other some embodiments", "in yet some embodiments", and the like appearing in various places in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "comprise", "include", "have", and their variants mean "including but not limited to", unless otherwise specifically emphasized.

[0033] It should be understood that the size of the serial number of each step in the following embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0034] In order to illustrate the technical solutions of the present application, the following will be illustrated by specific examples.

[0035] An embodiment of the present application provides a method for processing a seat staff, which can be applied in an application environment as shown in the figure. Figure 1 The client includes but is not limited to a palm computer, a desktop computer, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a cloud computer device, a personal digital assistant (PDA) and the like computer device. The server can be an independent server, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content distribution networks (CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

[0036] Please refer to Figure 2 An embodiment of the present application provides a method for processing a seat staff, which includes the following steps.

[0037] S100, historical work performance data of all seat staff in a seat queue is acquired;

[0038] S200, the acquired historical work performance data is preprocessed by cleaning and normalization, and feature extraction is performed on the preprocessed historical work performance data to obtain a feature vector;

[0039] S300, a call volume prediction result of the seat queue is generated by a pre-trained call volume prediction model based on the feature vector;

[0040] S400, a task allocation strategy is determined according to the call volume prediction result and current working states of the seat staff in the seat queue, and tasks are automatically assigned to the seat staff according to the task allocation strategy.

[0041] In specific implementation, the agent processing method of the embodiment realizes the optimization of agent task allocation and personnel management through a series of systematic steps. First, by obtaining the historical job performance data of all agents in the agent queue (S100), a rich data basis is provided for subsequent analysis. These data are preprocessed through cleaning and normalization and feature extraction (S200) to convert them into high-quality feature vectors that can accurately reflect the work records, work capacity and characteristics of the agents. Based on these feature vectors, a pre-trained call volume prediction model is used to generate call volume prediction results (S300), enabling managers to anticipate the future trends and distribution of call volume. Finally, in combination with the call volume prediction results and the current work status of the agents, a precise task allocation strategy is determined, and the agents are automatically assigned call tasks accordingly (S400). This method not only allocates tasks reasonably according to the actual ability and real-time state of the agents, avoiding uneven task allocation and resource waste, but also predicts call volume in advance, making personnel arrangements and resource allocation in advance, thereby improving the overall operation efficiency and service quality of the call center, enhancing customer satisfaction and enterprise competitiveness.

[0042] Understandably, the agent processing method provided by the embodiment of the present application can be applied to the agent processing scenarios related to the medical health field. The following is a specific example:

[0043] Suppose a large hospital's call center is responsible for handling patient appointment registration, consultation, postoperative follow-up and other businesses. After the call center introduces the agent processing method of the present application, the specific implementation process is as follows:

[0044] 1. Obtain historical job performance data (S100):

[0045] The call center system collects various types of medical consultation task data handled by the agents in the past period of time, including consultation types (such as appointment registration, disease consultation, drug consultation, etc.), processing time, patient satisfaction score, consultation results (such as successful appointment, answering questions, etc.), and the professional background of the agents (such as whether they have medical nursing knowledge, specific department consultation experience, etc.).

[0046] 2. Data preprocessing and feature extraction (S200):

[0047] The collected data are cleaned to remove invalid or erroneous records, such as duplicate data, data missing key information, etc. Then, normalization processing is performed to convert data of different dimensions to the same range, facilitating subsequent analysis.

[0048] Extract feature vectors, including the average processing time of the staff, the average patient satisfaction, the proportion of the number of each type of consultation, professional background labels (such as "nursing experience" and "familiar with internal medicine consultation"), and the like. These feature vectors can comprehensively reflect the work ability and characteristics of the staff.

[0049] 3. Call volume prediction (S300):

[0050] Using the pre-trained call volume prediction model, combined with the historical call data of the hospital and the recent business dynamics (such as the seasonal high incidence of diseases, the opening of new departments in the hospital, etc.), the call volume in different time periods in the future few days and the call volume proportion of each type of consultation are predicted. For example, it is predicted that during the high incidence of influenza, the call volume of disease consultation and appointment registration will increase significantly.

[0051] 4. Task allocation and personnel management (S400):

[0052] According to the call volume prediction results and the real-time working state of the staff (such as whether the current is idle, the number of tasks handled, etc.), the task allocation strategy is determined. The system will preferentially allocate disease consultation tasks to staff with medical care background and currently idle, and allocate appointment registration tasks to staff with high efficiency in handling this type of consultation.

[0053] At the same time, the task processing progress and patient feedback of the staff are monitored in real time. If it is found that the patient satisfaction handled by a staff is low, the system will timely adjust the task allocation, allocate subsequent tasks to other suitable staff, and provide targeted training suggestions for the staff, such as strengthening medical knowledge learning or communication skill training. In this way, the call center can efficiently and accurately handle various medical consultation tasks, improve patient satisfaction and the service quality of the hospital.

[0054] It can be understood that the staff processing method provided by the embodiment of the application can also be applied to the staff processing scene related to the field of financial technology. The following is a specific example:

[0055] Suppose the call center of a certain financial technology company is responsible for handling customer consultations and handling of online payment, credit application, investment consultation and other businesses. After the company adopts the staff processing method of the application, the specific implementation process is as follows:

[0056] 1. Obtain historical work performance data (S100):

[0057] The call center system collects data on various types of financial business consultation tasks handled by the agents in the past period of time, including consultation types (such as online payment problem consultation, credit application process consultation, investment product consultation, etc.), processing time, customer satisfaction score, consultation result (such as successfully solving problems, guiding customers to complete the application, etc.), and the business expertise of the agents (such as familiarity with credit business, good at investment consultation, etc.).

[0058] 2. Data preprocessing and feature extraction (S200):

[0059] The collected data is cleaned to remove invalid or erroneous records, such as duplicate data, data missing key information, etc. Then normalization processing is performed to convert data of different dimensions to the same range, facilitating subsequent analysis.

[0060] Feature vectors are extracted, including the average processing time of the agents, the mean customer satisfaction, the proportion of the number of times of processing each type of consultation, and the business expertise labels (such as "credit expert" and "investment consultant"), etc. These feature vectors can comprehensively reflect the work ability and characteristics of the agents.

[0061] 3. Call volume prediction (S300):

[0062] Using a pre-trained call volume prediction model, combined with the historical call data of the financial technology company and recent business dynamics (such as during promotional activities, new product launches, etc.), the call volume in different time periods in the future few days is predicted, as well as the call volume proportion of each type of consultation. For example, it is predicted that during the period of new product launch, the call volume of investment consultation and credit application consultation will significantly increase.

[0063] 4. Task allocation and personnel management (S400):

[0064] According to the call volume prediction results and the real-time working status of the agents (such as whether they are currently idle, the number of tasks they have processed, etc.), the task allocation strategy is determined. The system will preferentially allocate investment consultation tasks to agents with investment business expertise and who are currently idle, and allocate credit application consultation tasks to agents who are efficient in processing this type of consultation and familiar with the credit business process.

[0065] At the same time, the task processing progress and customer feedback of the agents are monitored in real time. If it is found that the customer satisfaction of a certain agent is low, the system will timely adjust the task allocation, allocate subsequent tasks to other suitable agents, and provide targeted training suggestions for the agent, such as strengthening business knowledge learning or communication skills training, etc. In this way, the call center can efficiently and accurately handle various types of financial business consultation tasks, improve customer satisfaction and the service quality of the company, effectively prevent business risks, and enhance the market competitiveness of the company.

[0066] Through the specific application examples of the two fields, it can be seen that the agent processing method can flexibly perform task allocation and personnel management according to the business characteristics and needs of different fields, significantly improves the operation efficiency and service quality of the call center, and has wide applicability and practicality.

[0067] Further, in an embodiment, the agent processing method, wherein the obtaining the historical job performance data of all the agents in the agent queue comprises the steps of:

[0068] Using a data source identification device, a plurality of data sources related to the stored data of the agent queue are determined;

[0069] Using a data bridging technology, a connection with each of the data sources is established;

[0070] From each of the data sources, historical job performance data related to all the agents in the agent queue is obtained.

[0071] In specific implementation, the specific implementation process of the steps of the present embodiment is approximately as follows:

[0072] 1. Data source identification

[0073] Step 1.1: Data source scanning

[0074] Using a data source identification device, the information system of the enterprise is scanned to identify a plurality of data sources related to the agent queue. These data sources can include a customer relationship management system (CRM), a call center management system (CCM), an agent operation log system, a customer feedback system, etc.

[0075] The identification device quickly locates the data sources that can contain historical job performance data of the agents through pre-set keywords (such as “agent”, “job record”, “customer feedback”, etc.) and data structure characteristics (such as database table name, field name, etc.).

[0076] Step 1.2: Data source verification

[0077] The identified data sources are verified to ensure that they contain historical job performance data of the agents. The verification process can be completed by checking whether the key fields (such as agent ID, job time, job type, customer satisfaction, etc.) exist in the data sources.

[0078] For data sources that do not meet the requirements, relevant information is recorded and excluded from the subsequent data acquisition process.

[0079] 2. Data bridging

[0080] Step 2.1: Establishing a connection

[0081] Connect to each validated data source using data bridging techniques. Data bridging techniques can include API interfaces, database connection tools (such as ODBC, JDBC), or middleware (such as ETL tools).

[0082] Configure appropriate connection parameters for each data source, such as server address, port number, username, password, etc., and perform connection testing to ensure stability.

[0083] Step 2.2: Permission Configuration

[0084] After establishing a connection, configure data access permissions to ensure safe access to data in the data source. Permission configuration can be set according to the enterprise's security policy, for example, only allow access to specific fields or data within a specific time period.

[0085] Adopt dynamic authentication mechanisms to verify the legitimacy of access requests in real time, preventing unauthorized access and data leakage.

[0086] 3. Data Acquisition

[0087] Step 3.1: Data Extraction

[0088] According to the pre-set data extraction rules, extract historical performance data of all agents in the agent queue from each data source. Data extraction rules can include extraction time range, data fields, etc.

[0089] For example, extract customer satisfaction scores and customer feedback records of agents from the CRM system; extract data such as work time, work type, and processing duration of agents from the call center management system.

[0090] Step 3.2: Data Integration

[0091] Integrate data extracted from multiple data sources to build a unified data set. The integration process needs to address field differences between data sources, inconsistent data formats, etc.

[0092] For example, use field mapping algorithms to match and unify fields representing the same information in different data sources; use data interpolation methods to fill in missing data caused by data source differences.

[0093] Step 3.3: Data Storage

[0094] Store the integrated data in a central database or data warehouse for subsequent data preprocessing and analysis. The storage process needs to ensure data integrity and consistency.

[0095] For example, data is stored in a relational database, indexed by fields such as agent ID, work time, etc., to improve data query efficiency.

[0096] Through the above process, the embodiment can efficiently obtain the historical work performance data of all current agents in the agent queue, providing a solid foundation for subsequent data preprocessing and analysis.

[0097] Further, in one embodiment, the agent processing method, wherein the historical work performance data is preprocessed by cleaning and normalizing, and the historical work performance data after preprocessing is feature extracted to obtain a feature vector, specifically including steps of:

[0098] Cleaning the historical work performance data by data denoising, missing value completion and similar value merging;

[0099] Normalizing the historical work performance data after cleaning, and extracting text features, numerical features, category features and time series features of the historical work performance data after normalization;

[0100] Fusing the extracted text features, numerical features, category features and time series features to obtain a feature vector of the historical work performance data.

[0101] In specific implementation, the specific implementation process of the steps of the embodiment is as follows:

[0102] 1. Data cleaning

[0103] Step 1.1: Data denoising

[0104] Use a deep learning algorithm (such as an autoencoder) to denoise the historical work performance data. By training an autoencoder model, the data is compressed into a low-dimensional space and then reconstructed, identifying and removing noise points that differ greatly from the original data, thereby improving data quality.

[0105] Step 1.2: Missing value completion

[0106] Use an intelligent filling algorithm based on machine learning to fill different types of missing data. For numerical data, use a time series prediction model (such as ARIMA or LSTM) to predict missing values based on data at adjacent time points; for categorical data, use a classification algorithm (such as decision tree or random forest) to predict missing category labels based on other relevant features, ensuring data integrity.

[0107] Step 1.3: Similar value merging

[0108] Similar records in the data are detected through hashing algorithms and similarity calculation techniques such as cosine similarity or Jaccard similarity. For records with a similarity exceeding a set threshold, an intelligent merging strategy is employed to retain key information and remove redundant data, avoiding the impact of data duplication on subsequent analysis.

[0109] 2. Data normalization and feature extraction

[0110] Step 2.1: Numerical normalization and feature extraction

[0111] After cleaning the numerical data, normalization processing is performed using Z-score standardization or Min-Max normalization methods to convert the data to a unified range, eliminating the influence of different dimensions and orders of magnitude on subsequent analysis. At the same time, numerical data is directly extracted as numerical features, or mathematical transformations (such as logarithmic transformation) are performed to improve data distribution and better reflect the internal rules of the data.

[0112] Step 2.2: Time format unification and feature extraction

[0113] Time data is converted to a unified time format (such as Unix timestamp) to facilitate subsequent time series analysis. Further, date, hour, and week information is extracted from time data as time series features, or statistical features (such as mean and variance) are calculated to more comprehensively characterize data characteristics in the time dimension.

[0114] Step 2.3: Classification encoding and feature extraction

[0115] For categorical data, One-Hot Encoding or Label Encoding is used for processing, converting it to numerical form so that machine learning models can recognize and process it. For example, skill labels of agents, job types, and other categorical features are encoded to generate corresponding category features, providing rich semantic information for subsequent analysis.

[0116] Step 2.4: Text feature extraction

[0117] For text content in historical job performance data (such as customer feedback, job descriptions, etc.), TF-IDF or Word2Vec methods are used to extract text features. By calculating term frequency-inverse document frequency (TF-IDF) or generating word embedding vectors (Word2Vec), text data is converted to a numerical feature vector, thereby transforming unstructured text information into structured data that can be used for analysis.

[0118] 3. Feature fusion

[0119] The extracted text features, numerical features, categorical features, and time series features are fused to obtain a complete feature vector of historical job performance data. Different types of features are combined into a unified feature vector through methods such as feature concatenation or feature weighted sum, providing high-quality input data for subsequent call volume prediction and task allocation strategy formulation, ensuring that the model can fully learn and utilize various information in the data, thereby achieving more accurate task allocation and personnel management.

[0120] Further, in one embodiment, the agent handling method, wherein the call volume prediction result of the agent queue is generated based on the feature vector through a pre-trained call volume prediction model, specifically comprising the steps of:

[0121] loading a pre-trained call volume prediction model;

[0122] matching the feature vector with the input format of the call volume prediction model;

[0123] if the format matching is successful, inputting the feature vector into the call volume prediction model to generate the call volume prediction result of the agent queue.

[0124] Further, the agent handling method, wherein after matching the feature vector with the input format of the call volume prediction model, it further specifically comprises the steps of:

[0125] if the format matching fails, normalizing the feature vector;

[0126] matching the normalized feature vector with the input format of the call volume prediction model again;

[0127] until the format matching is successful, inputting the feature vector into the call volume prediction model to generate the call volume prediction result of the agent queue.

[0128] In specific implementation, the specific implementation process of the steps of the present embodiment is approximately as follows:

[0129] 1. loading a pre-trained call volume prediction model

[0130] Step 1.1: Model loading

[0131] Load the pre-trained call volume prediction model from the model storage system. The model can be trained based on machine learning algorithms (such as random forest, support vector machine) or deep learning algorithms (such as LSTM, Transformer).

[0132] Ensure the integrity and consistency of the model file, such as verifying the integrity of the model file through hash check or digital signature, to prevent the model file from being tampered with or damaged during loading.

[0133] 2. Feature vector matches the model input format

[0134] Step 2.1: Format check

[0135] Match the extracted feature vector with the input format of the call volume prediction model. Check whether the dimension, data type and order of the feature vector are consistent with the input requirements of the model.

[0136] For example, the model may require the dimension of the input feature vector to be 100, the data type to be floating point, and the order of the features to be [numeric features, text features, category features, time series features].

[0137] 3. Format matching success, generate prediction results

[0138] Step 3.1: Input model

[0139] If the format matches successfully, the feature vector is directly input into the call volume prediction model.

[0140] The model will generate call volume prediction results for the agent queue according to the input feature vector, which can include call volume, call type distribution and other information in the future time period.

[0141] 4. Format matching fails, standardization processing

[0142] Step 4.1: Standardization processing

[0143] If the format matching fails, the feature vector is standardized. Standardization processing includes:

[0144] Data type conversion: convert the data type in the feature vector to the type required by the model (such as converting integer to floating point).

[0145] Dimension adjustment: if the dimension of the feature vector is inconsistent with the model input, the dimension can be adjusted by padding or cropping.

[0146] Order adjustment: if the order of the feature vector is inconsistent with the model input, the order of the feature vector can be rearranged.

[0147] Step 4.2: Re-match

[0148] Match the feature vector after standardization with the input format of the call volume prediction model again, check whether it meets the input requirements of the model.

[0149] 5. Repeat processing until matching succeeds

[0150] Step 5.1: Cycle processing

[0151] If the re-matching still fails, repeat steps 4.1 and 4.2 to further standardize the feature vector until the format matching is successful.

[0152] Step 5.2: Generate prediction results

[0153] Once the format matching is successful, input the standardized feature vector into the call volume prediction model to generate call volume prediction results for the agent queue.

[0154] Through the above process, the embodiment can ensure that the feature vector is consistent with the input format of the call volume prediction model, thereby generating accurate call volume prediction results to provide reliable data support for subsequent task allocation and personnel management.

[0155] Further, in one embodiment, the agent personnel processing method, wherein the task allocation strategy is determined according to the call volume prediction results and the current working state of each agent personnel in the agent queue, and tasks are automatically assigned to each agent personnel according to the task allocation strategy, specifically including steps of:

[0156] Obtaining the current working state of each agent personnel in the agent queue, sorting the target agent personnel in the idle state in the agent queue according to a preset sorting rule to generate an idle agent list;

[0157] According to the idle agent list and the call volume prediction results, a task allocation strategy is determined, and tasks are automatically assigned to each agent personnel according to the task allocation strategy.

[0158] Further, the agent personnel processing method, wherein the current working state of each agent personnel in the agent queue is obtained, and the target agent personnel in the idle state in the agent queue is sorted according to a preset sorting rule to generate an idle agent list, specifically including steps of:

[0159] Obtaining the current working state of each agent personnel in the agent queue, determining the target agent personnel in the idle state in the agent queue;

[0160] According to the skill label of each target agent personnel, an ability score is assigned to each target agent personnel;

[0161] Each target agent personnel is sorted according to the ability score to generate an idle agent list.

[0162] In implementation, the specific implementation process of the steps of this embodiment is approximately as follows:

[0163] 1. Obtain the current working status of the agent personnel

[0164] Step 1.1: Real-time monitoring of agent status

[0165] Real-time acquisition of the current working status of each agent personnel in the agent queue through the agent management system. The working status includes but is not limited to: idle, busy, online duration, current processing customer type, historical customer satisfaction score, etc.

[0166] Utilize Internet of Things (IoT) technology or real-time data streaming technology (such as Apache Kafka) to ensure that the working status of the agent personnel can be updated in real time and transmitted to the call center system.

[0167] Step 1.2: Screening idle agent personnel

[0168] Screen the target agent personnel currently in idle state from all agent personnel. The idle state can be defined as the agent personnel currently having no ongoing tasks or having a workload lower than a set threshold.

[0169] Record the screened idle agent personnel for subsequent sorting and task allocation.

[0170] 2. Assign a capability score to each idle agent personnel

[0171] Step 2.1: Extract skill tags

[0172] According to the skill tags (such as language proficiency, business expertise, processing speed, historical performance, etc.) of each idle agent personnel, extract relevant information. Skill tags can be obtained from historical job performance data of agent personnel or updated through regular skill assessment tests.

[0173] For example, skill tags can include:

[0174] Language proficiency: Whether proficient in multiple languages.

[0175] Business expertise: Whether familiar with specific business areas (such as medical consultation, financial credit, etc.).

[0176] Processing speed: Average time to process tasks.

[0177] Historical performance: Customer satisfaction score, task success rate, etc.

[0178] Step 2.2: Calculate capability score

[0179] According to the pre-set scoring rules, assign a capability score to each idle agent personnel.

[0180] The scoring rules can be assigned to different skill labels based on different weights.

[0181] 3. Generate a list of idle agents

[0182] Step 3.1: Sort by ability score

[0183] Sort all idle agents by their ability scores from high to low. Agents with higher ability scores will be given priority in task assignment.

[0184] Generate a list of idle agents, which will serve as the basis for task assignment.

[0185] 4. Determine the task assignment strategy

[0186] Step 4.1: Analyze call volume prediction results

[0187] Based on the prediction results generated by the call volume prediction model, analyze the call volume trends, peak periods, and call type distribution in the future time period. The prediction results should include detailed information such as hourly call volume, call type proportion, etc.

[0188] For example, the prediction results show that the number of consultation calls will increase significantly, while the number of complaint calls is relatively small in a certain time period.

[0189] Step 4.2: Develop a task assignment strategy

[0190] Based on the list of idle agents and the call volume prediction results, develop a task assignment strategy. The strategy should include:

[0191] Task priority: Assign priority to each call task based on call type and urgency.

[0192] Task assignment rules: Assign call tasks to the most suitable agents based on their ability scores and skill labels.

[0193] Resource optimization: Reasonably arrange the workload of agents according to the peak and trough periods of call volume, to avoid excessive fatigue or resource idleness.

[0194] For example, preferentially assign high-value customer or emergency call tasks to agents with high ability scores to ensure that critical tasks can be handled in a timely manner.

[0195] 5. Assign tasks to agents

[0196] Step 5.1: Real-time task pushing

[0197] According to the task allocation strategy, real-time task pushing is performed to idle agents through the agent management system. The task pushing should include call types, estimated processing time, priority, and other key information to ensure that the agents can quickly understand the task requirements.

[0198] After the agents receive the tasks, the timestamp of task allocation and the response status of the agents (such as acceptance or rejection) are recorded.

[0199] Step 5.2: Dynamic adjustment and optimization

[0200] Real-time monitoring of task processing progress and call volume changes, dynamic adjustment of task allocation strategy according to actual situation. For example, if the task processing time of an agent is too long or the call volume prediction result changes, the system will automatically adjust the task allocation to ensure the optimality of resource utilization.

[0201] If an agent rejects a task or does not respond within the specified time, the system automatically reassigns the task to the next suitable agent.

[0202] Step 5.3: Performance evaluation and feedback

[0203] Real-time monitoring of the task execution of the agents, including task processing time, customer satisfaction score and other key indicators. Real-time performance reports of the agents are generated using big data analysis technology.

[0204] According to the performance evaluation results, personalized performance feedback and training suggestions are provided to the agents. For example, if an agent performs poorly in handling a certain type of call, the system can recommend relevant training courses or provide real-time guidance.

[0205] Through the above process, the embodiment can achieve efficient task allocation and management of agents, ensuring that the call center can reasonably utilize agent resources and improve overall operational efficiency and service quality when facing dynamic changes in call volume.

[0206] As can be seen from the above method embodiment, the agent processing method provided by the present application comprises: obtaining historical job performance data of all agents in the agent queue; preprocessing the obtained historical job performance data by cleaning and normalizing, and extracting features from the preprocessed historical job performance data to obtain a feature vector; generating a call volume prediction result of the agent queue based on the feature vector through a pre-trained call volume prediction model; determining a task allocation strategy according to the call volume prediction result and the current working state of each agent in the agent queue, and automatically assigning tasks to each agent according to the task allocation strategy. In this way, the task allocation and personnel management of the agents can be optimized through the method of the present application.

[0207] It should be understood that although the present application provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-creative work, and these operation steps are not necessarily performed in the order of the embodiments or flowcharts. The order of steps listed in the embodiments or flowcharts is only one way of executing the steps among many steps and does not represent the only execution order. It should be noted that there is not necessarily a certain order between the above steps. Those of ordinary skill in the art can understand from the description of the embodiments of the present invention that in different embodiments, the above steps may have different execution orders, that is, they may be executed in parallel, or they may be executed in an interchangeable manner, etc. Moreover, at least a portion of the steps in the embodiments or flowcharts may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but may be executed at different times. The execution order of these sub-steps or stages is not necessarily to be performed in sequence, but may be executed in turn, alternately or synchronously with other steps or at least a portion of the sub-steps or stages of other steps.

[0208] Based on the above method embodiment, please refer to Figure 3 Another embodiment of the present invention further provides an agent processing device, wherein the device includes:

[0209] The data acquisition module 11 is used to obtain the historical performance data of all agents currently in the agent queue;

[0210] A feature extraction module 12 is configured to perform cleaning and normalization preprocessing on the acquired historical operation performance data, and perform feature extraction on the preprocessed historical operation performance data to obtain a feature vector;

[0211] A result generating module 13 is configured to generate a call volume prediction result of the agent queue based on the feature vector using a pre-trained call volume prediction model;

[0212] The agent management module 14 is configured to determine a task allocation strategy based on the call volume prediction result and the current working status of each agent in the agent queue, and automatically assign tasks to each agent according to the task allocation strategy.

[0213] Furthermore, in one embodiment, the agent processing device, wherein the step of obtaining historical work performance data of all agents currently in the agent queue, specifically includes:

[0214] using a data source identification device to determine a plurality of data sources associated with stored data of the agent queue;

[0215] Using data bridging technology to establish connections with each of the data sources;

[0216] obtaining historical job performance data related to all current agents in the agent queue from each of the data sources.

[0217] Further, in an embodiment, the agent processing apparatus, wherein the historical job performance data obtained is preprocessed by cleaning and normalizing, and feature extraction is performed on the preprocessed historical job performance data to obtain a feature vector, specifically comprising:

[0218] cleaning the historical job performance data by data denoising, missing value completion, and similar value merging;

[0219] normalizing the historical job performance data after cleaning, and extracting text features, numerical features, category features, and time series features of the historical job performance data after normalization;

[0220] fusing the extracted text features, numerical features, category features, and time series features to obtain a feature vector of the historical job performance data.

[0221] Further, in an embodiment, the agent processing apparatus, wherein the call volume prediction result of the agent queue is generated based on the feature vector by a pre-trained call volume prediction model, specifically comprising:

[0222] loading the pre-trained call volume prediction model;

[0223] matching the feature vector with an input format of the call volume prediction model;

[0224] if the format matching is successful, inputting the feature vector into the call volume prediction model to generate the call volume prediction result of the agent queue.

[0225] Further, the agent processing apparatus, wherein after matching the feature vector with the input format of the call volume prediction model, further specifically comprising:

[0226] if the format matching fails, performing standardization processing on the feature vector;

[0227] matching the feature vector after standardization processing with the input format of the call volume prediction model again;

[0228] until the format matching is successful, inputting the feature vector into the call volume prediction model to generate the call volume prediction result of the agent queue.

[0229] Further, in one embodiment, the agent processing apparatus, wherein the determining the task allocation strategy according to the call volume prediction result and the current working state of each of the agents in the agent queue, and automatically assigning tasks to each of the agents according to the task allocation strategy, specifically comprises:

[0230] obtaining the current working state of each of the agents in the agent queue, sorting the target agents in the idle state in the agent queue according to a preset sorting rule to generate an idle agent list;

[0231] determining the task allocation strategy according to the idle agent list and the call volume prediction result, and automatically assigning tasks to each of the agents according to the task allocation strategy.

[0232] Further, the agent processing apparatus, wherein the obtaining the current working state of each of the agents in the agent queue, sorting the target agents in the idle state in the agent queue according to a preset sorting rule to generate an idle agent list, specifically comprises:

[0233] obtaining the current working state of each of the agents in the agent queue, and determining the target agents in the idle state in the agent queue;

[0234] assigning an ability score to each of the target agents according to the skill label of each of the target agents;

[0235] sorting each of the target agents according to the ability score to generate an idle agent list.

[0236] It should be noted that the information interaction, execution process and the like between the above modules in the device embodiment of the application are based on the same concept as the method embodiment of the application, and the specific functions and the technical effects brought by the same can be referred to the method embodiment part, which will not be described here.

[0237] Based on the above method embodiment, another embodiment of the application further provides a computer device, which can be a server, and the internal structure diagram thereof can be as shown in Figure 4As shown in the structural schematic diagram. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with the external terminal through the network connection. The computer program is executed by the processor to realize the function or step of the agent processing method server side in any one of the above method embodiments.

[0238] Based on the above method embodiments, another embodiment of the present application further provides a computer device which can be a client, and its internal structure diagram can be as shown in the structural schematic diagram. Figure 5 The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with the external terminal through the network connection. The computer program is executed by the processor to realize the function or step of the agent processing method client side in any one of the above method embodiments.

[0239] Those skilled in the art can understand that, Figure 4 As shown in the structural schematic diagram, Figure 5 The structural schematic diagram shown in the above embodiment is only a schematic diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. Specifically, the computer device can include more components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0240] Among them, the processor can be a CPU, and the processor can also be other general-purpose processors, digital signal processors (Digital Signal Processors, DSPs), application specific integrated circuits (Application Specific Integrated Circuits, ASICs), ready programmable gate arrays (Field-Programmable Gate Arrays, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0241] The memory includes a readable storage medium, an internal memory, etc., wherein the internal memory can be a memory of the computer device, and the internal memory provides an environment for running the operating system and the computer readable instructions in the readable storage medium. The readable storage medium can be a hard disk of the computer device, and in other embodiments, can also be an external storage device of the computer device, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory can include both an internal storage unit of the computer device and an external storage device. The memory is used to store an operating system, an application program, a boot loader, data, and other programs, such as program codes of computer programs, etc. The memory can also be used to temporarily store data that has been output or will be output.

[0242] Based on the above method embodiments, another embodiment of the present application further provides a computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the agent processing method in any one of the above method embodiments. The computer readable storage medium can be non-volatile or volatile.

[0243] It should be noted that the functions or steps that the computer readable storage medium or the computer device can achieve and the technical effects brought by the functions / steps can be referred to the related description in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0244] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, storage, database or other medium used in each embodiment provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc. The disclosed memory components or memories of the operating environment described herein are intended to include one or more of these and / or any other suitable type of memory.

[0245] Those skilled in the art can clearly understand that, for the convenience and brevity of description, in the device embodiments of the present application, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the above-described functions. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or software. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the unit and module in the above-mentioned device can refer to the corresponding process in the above-mentioned method embodiments, which will not be repeated here. If the integrated unit is realized in the form of software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium.

[0246] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0247] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / computer device and method can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely schematic. The division of the modules or units is merely a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0248] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0249] It should be noted that if non-company software tools or components appear in the embodiments of the present application, they are only used for example introduction and do not represent actual use. The above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for handling seat personnel, characterized in that: include: Get the historical performance data of all agents currently in the agent queue; Performing preprocessing of cleaning and normalization on the acquired historical operation performance data, and performing feature extraction on the preprocessed historical operation performance data to obtain a feature vector; Based on the feature vector, generating a call volume prediction result for the agent queue through a pre-trained call volume prediction model; A task allocation strategy is determined based on the call volume prediction result and the current working status of each agent in the agent queue, and tasks are automatically assigned to each agent according to the task allocation strategy.

2. The agent handling method according to claim 1, characterized in that: The acquisition of historical performance data of all agents currently in the agent queue includes: using a data source identification device to determine a plurality of data sources associated with stored data of the agent queue; Using data bridging technology to establish connections with each of the data sources; The historical operation performance data related to all current agents in the agent queue is obtained from each of the data sources.

3. The agent handling method according to claim 1, characterized in that: The preprocessing of cleaning and normalizing the acquired historical operation performance data, and extracting features from the preprocessed historical operation performance data to obtain a feature vector, includes: Performing cleaning processing on the historical operation performance data by removing noise, filling in missing values, and merging similar values; Normalizing the cleaned historical operation performance data, and extracting text features, numerical features, category features, and time series features of the normalized historical operation performance data; The extracted text features, the numerical features, the category features, and the time series features are fused to obtain a feature vector of the historical job performance data.

4. The agent handling method according to claim 1, characterized in that: Generating a call volume prediction result of the agent queue based on the feature vector using a pre-trained call volume prediction model includes: Load the pre-trained call volume prediction model; Matching the feature vector with an input format of the call volume prediction model; If the format matches successfully, the feature vector is input into the call volume prediction model to generate a call volume prediction result for the agent queue.

5. The agent processing method according to claim 4, characterized in that: After matching the feature vector with the input format of the call volume prediction model, the method further includes: If the format matching fails, the feature vector is normalized; Matching the normalized feature vector again with the input format of the call volume prediction model; When the format is successfully matched, the feature vector is input into the call volume prediction model to generate a call volume prediction result for the agent queue.

6. The agent handling method according to claim 1, characterized in that: Determining a task allocation strategy based on the call volume prediction result and the current working status of each agent in the agent queue, and automatically assigning tasks to each agent according to the task allocation strategy, includes: Obtaining the current working status of each agent in the agent queue, sorting the target agents currently in the agent queue in an idle state according to a preset sorting rule, and generating an idle agent list; A task allocation strategy is determined based on the idle agent list and the call volume prediction result, and tasks are automatically assigned to each agent according to the task allocation strategy.

7. The agent handling method according to claim 6, characterized in that: The step of obtaining the current working status of each agent in the agent queue, sorting the target agents currently in an idle state in the agent queue according to a preset sorting rule, and generating an idle agent list includes: Obtaining the current working status of each agent in the agent queue, and determining a target agent in the agent queue who is currently idle; Assigning a capability score to each target agent based on the skill tag of each target agent; The target agents are sorted according to their ability scores to generate a list of available agents.

8. A seat personnel processing device, characterized in that: include: The data acquisition module is used to obtain the historical performance data of all agents currently in the agent queue; a feature extraction module, configured to perform cleaning and normalization preprocessing on the acquired historical operation performance data, and perform feature extraction on the preprocessed historical operation performance data to obtain a feature vector; A result generation module, configured to generate a call volume prediction result for the agent queue based on the feature vector using a pre-trained call volume prediction model; The agent management module is used to determine a task allocation strategy based on the call volume prediction result and the current working status of each agent in the agent queue, and automatically assign tasks to each agent according to the task allocation strategy.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the agent processing method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the agent processing method according to any one of claims 1 to 7 is implemented.

Images