Enterprise customer operation and maintenance management system based on micro service

Abstract

The invention discloses an enterprise customer operation and maintenance management system based on micro-service, and relates to the technical field of operation and maintenance management, and the system specifically comprises a recessive dynamic data collection module, a recessive dynamic feature extraction module and a customer priority scenarized calculation module. The method comprises the following steps: setting different data acquisition periods according to distribution characteristics of system operation frequency data, collecting implicit dynamic data from multiple dimensions in each acquisition period, and extracting characteristic indexes of the implicit dynamic data based on historical operation and maintenance data correlation and set parameters; analyzing interaction association features set by dynamic scheduling feature indexes and customer basic attribute comprehensive values, determining urgency association factors and value potential cooperation factors, constructing a nonlinear mapping model to quantify enterprise customer priorities in combination with an enterprise real-time operation and maintenance scene, dynamically adjusting an operation and maintenance strategy in combination with a resource state and a service period, and establishing an enterprise real-time operation and maintenance strategy. And the balance between the short-term operation and maintenance efficiency and the long-term customer value is realized.

Description

Technical Field

[0001] This invention belongs to the field of operation and maintenance management technology, specifically relating to a microservice-based enterprise customer operation and maintenance management system. Background Technology

[0002] Against the backdrop of accelerated digital transformation, enterprise customers have significantly increased their demands for real-time, accurate, and personalized operation and maintenance (O&M) services. Especially with the widespread adoption of microservice architectures, customer O&M scenarios exhibit three main characteristics: increased customer scale, more complex business logic, and dynamic resource fluctuations. However, traditional enterprise customer O&M management solutions face numerous bottlenecks in terms of technical implementation and business adaptation, making it difficult to meet current O&M needs. Specific background technical issues are as follows: Traditional customer operations and maintenance data collection mechanisms are rigid, unable to balance real-time performance and resource efficiency, and fail to capture implicit risk signals. Traditional collection only focuses on explicit operations and maintenance signals (such as fault alarms and service response timeouts), ignoring implicit signals such as the number of repeated operations, the proportion of invalid operations, and non-peak resource requests. These signals can predict potential customer operations and maintenance needs in advance (such as repeated operations may indicate functional failures, and invalid operations may reflect usage confusion). However, due to the lack of collection, the transformation of operations and maintenance from passive response to proactive early warning is hindered, resulting in a high customer complaint rate.

[0003] In traditional operations and maintenance, customer priority judgment often relies on static basic data such as contract amount and cooperation period, or a single dynamic indicator (such as resource request frequency). This fails to achieve deep integration of basic data and dynamic signals, resulting in technical defects. This leads to linear processing of basic data, which deviates from the essence of operations and maintenance value. The basic data and dynamic signals become disconnected, and priority judgment becomes distorted. Therefore, there is an urgent need for a microservice-based enterprise customer operation and maintenance management system to solve the above problems. Summary of the Invention

[0004] The purpose of this invention is to provide a microservice-based enterprise customer operation and maintenance management system to solve the technical problems in the prior art, such as linear basic data processing, detachment from the essential value of operation and maintenance, disconnection between basic data and dynamic signals, and distortion of priority judgment.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A microservice-based enterprise customer operation and maintenance management system includes: The implicit dynamic data collection module is used to analyze the number of system operations of customers over a historical period. Based on the distribution characteristics of the number of system operations, it distinguishes different types of customers and sets different data collection cycles for different types of customers. In each collection cycle, it collects implicit dynamic data that affects the customer's operation and maintenance priority from multiple dimensions. The implicit dynamic data includes system behavior implicit data and interaction depth implicit data; The system behavior implicit data includes the number of repeated operations N1, the proportion of invalid operations R1, the frequency of resource requests during non-peak periods F1, and the duration of burst resource occupancy T1; The interaction depth implicit data includes the frequency of function extension consultations N2 and the effective viewing duration T2; The collected data is denoised and standardized. After standardization, the value range of all numerical values is between [0, M], and M is set according to actual requirements; The implicit dynamic feature extraction module sets parameters based on the relevance of historical operation and maintenance data to extract the feature indicators of implicit dynamic data, determines the standard values of the annual contract amount and the cooperation years, and联动修正生成基础属性综合值; The customer priority scenario calculation module is used to analyze the interaction correlation features set by the dynamic scheduling feature indicators and the comprehensive value of the customer's basic attributes, determine the urgency correlation factor and the value potential synergy factor, and construct a non-linear mapping model in combination with the enterprise's real-time operation and maintenance scenario to quantify the enterprise customer priority.

[0006] Further, analyze the system operation times data of customers within the historical time. According to the distribution characteristics of the system operation times data, different types of customers are distinguished and different data collection cycles are set for different types of customers. The specific method is as follows: Define the basic window with the duration T0 as the reference period, extract the system operation times of all customers within the historical T duration, set the high-active customer threshold X equal to the average value of the daily operation times of all customers plus the standard deviation, and set the low-active customer threshold Y equal to the average value of the daily operation times of all customers minus the standard deviation. Customers with a daily average system operation times > X are recorded as high-active customers, and the shrinking window T1 = T0 × a, a ∈ (0, 1), where a is the high-active window shrinking factor, and the setting of a depends on the data real-time requirement and system load. Customers with a daily average system operation times < Y are recorded as low-active customers, and the expanding window T2 = T0 × b, b > 1, where b is the low-active window expanding factor, and the setting of b is based on the data update inertia of low-active customers. When the daily average system operation times is within the interval [Y, X], the basic window is used.

[0007] Further, based on the relevance of historical operation and maintenance data, set parameters to extract the feature indicators of implicit dynamic data. The specific method is as follows: Set parameters Wh1 and Wh2 based on the relevance between N1, R1 and the customer complaint rate in historical operation and maintenance cases, and use the formula Calculate the operation health index; Set parameters Ws1 and Ws2 based on the correlation degree between F1, T1 and the expansion urgency in historical resource expansion data, and use the formula Calculate the computing resource demand intensity indicator; Set parameters We1 and We2 based on the correlation between N2, T2 and the growth of the final renewal amount in the historical cooperation expansion cases, and use the formula<0OO0033>To calculate the expansion potential indicator.

[0008] Furthermore, determine the standard value of the annual contract amount. The specific method is as follows: Divide the customer's annual contract amount into k levels, L1, L2,..., Lk. The level thresholds are set based on 1 / k of the average contract amount of enterprise customers. Based on the actual operation and maintenance value weight corresponding to the level contract amount and the enterprise resource input return ratio, set the corresponding level benchmark score for each level. The specific values are set according to historical data and actual requirements. Through training with historical operation and maintenance resource consumption data, determine the operation and maintenance resource consumption coefficient K1 for different types of customers; The standardized value of the annual contract amount corresponding to customer x is obtained from the formula Get, Represents the level benchmark score corresponding to customer x, Represents the operation and maintenance resource consumption coefficient corresponding to customer x, Represents the maximum level benchmark score.

[0009] Furthermore, determine the standardized value of the cooperation years. The specific method is as follows: Non-linearly divide the basic years score according to the cooperation years t. When t ≤ a1, set the basic years score St equal to A1, where A1 > 0. When a1 < t ≤ a2, set the basic years score When a2 < t, Where A2 = A1 + Among them And And Represent the change coefficients, which are determined by combining the relationship between the cooperation years and the enterprise operation and maintenance input-output ratio in historical data. a1 and a2 are the cooperation years thresholds, which are set according to historical data and actual requirements.

[0010] Based on the historical T duration data, determine the actual renewal times and the expected renewal times of the enterprise, and record the ratio of the actual renewal times to the expected renewal times as the renewal stability coefficient K2. Then, the standardized value of the cooperation years corresponding to customer x is calculated from the formula Obtained.

[0011] Furthermore,联动修正生成基础属性综合值,具体方法为: Customer basic attribute comprehensive value, where And Are used to balance And the cooperation years It should be noted that there is an unclear expression "联动修正生成基础属性综合值" in the original text. I have translated it as literally as possible. If there is a more accurate or specific expression, the translation can be adjusted accordingly.The weight in the customer's basic attributes is set according to the correlation between the two in terms of actual operation and maintenance resource consumption and customer retention, satisfying , is the weight of the linkage correction term, determined through business scenario analysis and historical data.

[0012] Furthermore, determine the urgency correlation factor and the value potential synergy factor. The specific method is as follows: Using the formula represents the urgency correlation factor F1, used to quantify the urgency of the customer's current operation and maintenance requirements, where is a positive integer, set according to actual requirements, is the urgency amplification coefficient, dynamically calibrated based on the customer's fault diffusion speed in historical data; Using the formula represents the value potential synergy factor F2, used to quantify the long-term operation and maintenance value of the customer. Among them, represents the synergy enhancement threshold. When E exceeds θ times of C, synergy enhancement is triggered, is the potential amplification coefficient, calibrated based on the actual cooperation expansion data in historical data. Both F1 and F2 are greater than 0.

[0013] Furthermore, combined with the enterprise's real-time operation and maintenance scenario, construct a non-linear mapping model to quantify the priority of enterprise customers. The specific method is as follows: Divide the enterprise operation and maintenance cycle into three categories: resource supply guarantee period, business expansion period, and regular operation and maintenance period. Each category of scenario generates the customer priority parameter P through and 's differential fusion logic; When the server load rate collected by the resource monitoring module ≥ L0, it is the resource supply guarantee period; When the server load rate < L0, but within the TA period after the enterprise's new business is launched or during the peak customer renewal period, it is the business expansion period. TA is set according to the actual situation; All periods other than the above two scenarios are recorded as the regular operation and maintenance period.

[0014] Furthermore, each category of scenario generates the customer priority parameter P through and 's differential fusion logic. The specific method is as follows: In the resource supply guarantee period, use the formula represents the customer priority parameter, is the bottom coefficient, set based on the reasonable resource occupancy ratio of high-F2 customers when resources are紧张; In the business expansion period, use the formula represents the customer priority parameter, is the filtering coefficient, set based on the proportion of urgent needs of high-potential customers in history; During routine maintenance, using formulas This indicates a customer priority parameter. Represents the equilibrium coefficient, based on historical data. and The distribution density of the difference is set.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention can predict potential customer operation and maintenance needs in advance by collecting implicit signals such as the number of repeated operations, the proportion of invalid operations, and non-peak resource requests. It adjusts feature weights based on the relevance of historical data and iteratively optimizes them through real-time feedback (resource over-limit rate, expansion conversion rate). This improves the matching degree between indicators such as operation health and resource demand intensity and actual business, and avoids feature distortion caused by fixed weights. 2. This invention, through standardized processing of basic data, combines the operation and maintenance resource consumption coefficients of customers in different industries and the average annual operation and maintenance resource ratio of customers at each amount level to ensure that customers with the same amount but different industries and different operation and maintenance costs can receive differentiated assessments. This avoids high operation and maintenance cost customers missing out on resource guarantees due to simple amount judgment. For the cooperation period, a basic score is divided through non-linear rules, and then combined with the renewal stability coefficient to clearly distinguish between long-term cooperation but unstable renewal and short-term cooperation but strong renewal intention. This effectively reduces the dissatisfaction of high-value customers due to the failure to handle real-time issues in a timely manner. 3. This invention focuses on the reverse synergy between low operational health and high resource demand by setting an urgency-related factor. When a customer's operational effectiveness is poor and resource requests are urgent, the urgency level will increase exponentially. This design can accurately identify high-risk customers with rapid fault propagation, improve the accuracy of priority identification for such customers, and effectively avoid the systemic risk of local anomalies escalating into full business interruption. At the same time, the value potential synergy factor captures the positive resonance between the customer's basic attributes and expansion potential. When the customer's basic value is high and expansion consultations are frequent with sufficient effective access time, a synergy enhancement mechanism will be triggered, amplifying its long-term operational value. The renewal amount growth rate of such high-growth potential customers can thus increase, locking in more long-term business value for the enterprise. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1Shows a module diagram of an enterprise customer operation and maintenance management system based on microservices according to the present invention; Figure 2 Shows a step diagram of a method for quantifying customer priorities according to the present invention; Figure 3 Shows a step diagram of a differential allocation method based on customer P values through a resource monitoring microservice. Specific embodiments

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] Such as Figure 1 、 Figure 2 、 Figure 3 A microservice-based enterprise customer operation and maintenance management system shown as follows specifically includes the following steps: An implicit dynamic data collection module is used to analyze the system operation times data of customers within a historical time period, distinguish different types of customers according to the distribution characteristics of the system operation times data, set different data collection cycles for different types of customers, and collect implicit dynamic data affecting customer operation and maintenance priorities from multiple dimensions within each collection cycle.

[0020] Define a basic window with a reference period of duration T0, extract the system operation times of all customers within a historical duration of T, set a high-active customer threshold X and a low-active customer threshold Y based on the distribution of historical customer operation data, where the high-active customer threshold X is equal to the average of the daily operation times of all customers plus one standard deviation, and the low-active customer threshold Y is equal to the average of the daily operation times of all customers minus one standard deviation. Customers with a daily system operation times > X are recorded as high-active customers, and a shrinking window T1 = T0 × a, a ∈ (0, 1), is used to collect data every T1 cycle to capture real-time behavior changes. a is the high-active window shrinkage factor, and the setting of a depends on the data real-time requirement and system load. The initial value is set as the ratio of T1 / T0, and then a is dynamically adjusted in combination with the monitoring data collection effect (such as signal capture rate). Customers with a daily system operation times < Y are recorded as low-active customers, and an extended window T2 = T0 × b, b > 1, is used to collect data every T2 cycle to reduce ineffective collection. b is the low-active window expansion factor, and the setting of b is based on the data update inertia of low-active customers. When the daily system operation times are within the interval [Y, X], data is collected every T0 cycle.

[0021] In each data collection cycle, implicit dynamic data is collected from multiple dimensions to cover key information that affects the customer's operation and maintenance priorities; Implicit dynamic data includes implicit data on system behavior and implicit data on interaction depth; The implicit data of system behavior includes the number of repeated operations N1, the proportion of invalid operations R1, the frequency of resource requests during off-peak periods F1, and the duration of sudden resource occupation T1; The number of repeated operations refers to the number of times a customer performs the same operation (such as submitting a form) consecutively within the same functional module but fails (e.g., if the customer clicks 3 times without submitting, then N1=3). Invalid operation percentage refers to the proportion of customer operations that do not produce business results (such as closing a page within 10 seconds after opening it, or not saving the content after entering it) out of the total number of operations. That is, R1 is equal to the ratio of invalid operation number to total operation number. Off-peak resource request frequency refers to the number of computing power or storage resource requests initiated by the customer system during off-peak hours (set according to historical data and actual conditions, such as 0:00-6:00 every day); The duration of sudden resource occupancy refers to the duration during which resource occupancy exceeds the customer's historical average threshold (set based on historical data and actual situation, such as exceeding twice the customer's historical average). The implicit data on interaction depth includes the frequency of function expansion consultations N2 and the effective viewing time T2; Among them, the frequency of function expansion consultations refers to the number of times customers search for and consult extended terms (such as new business integration and function adaptation) in addition to the preset keywords within the time window period; Effective browsing time refers to the cumulative time spent on a page that is greater than or equal to a preset time threshold. The time threshold is set based on historical data (for example, when a customer browses a manual, the cumulative time spent on the page is ≥30 seconds. If the page stays for 2 minutes, then T2=120 seconds; if the page stays for 20 seconds, then T2=0 seconds).

[0022] The collected data is denoised and standardized. After standardization, all values ​​are within the range of [0, M], where M is set according to actual needs.

[0023] The implicit dynamic feature extraction module extracts feature indicators of implicit dynamic data based on the correlation of historical operation and maintenance data, sets parameters to determine the standard values ​​of annual contract amount and cooperation period, and generates a comprehensive value of basic attributes through linkage correction.

[0024] The system receives the cleaned data output from step one via a standardized API, including system behavior signals (standardized values ​​of N1, R1, F1, and T1) and interaction depth signals (standardized values ​​of N2 and T2). At the same time, it retrieves its own stored customer basic data (including the standardized value of annual contract amount C1 and the standardized value of cooperation period C2).

[0025] Extract dynamic scheduling feature indicators (including operation health indicators, resource demand intensity indicators, and expansion potential indicators) by combining system behavior signals and interaction depth signals. Set parameters Wh1 and Wh2 based on the correlations between N1, R1 and the customer complaint rate in historical operation and maintenance cases (e.g., if N1 has a higher correlation with the complaint rate, Wh1 = 0.6, Wh2 = 0.4), and use the formula to calculate the operation health indicator; Set parameters Ws1 and Ws2 based on the correlation degrees between F1, T1 and the expansion urgency in historical resource expansion data (e.g., if T1 has a higher correlation with the resource overrun risk, then Ws2 = 0.6, Ws1 = 0.4), and adjust according to the resource utilization feedback later (if the resource overrun rate decreases after increasing Ws2, then keep the adjustment, otherwise readjust), and use the formula to calculate the resource demand intensity indicator; Set parameters We1 and We2 based on the correlations between N2, T2 and the growth of the final renewal amount in historical cooperation expansion cases (e.g., if N2 has a higher correlation with the renewal growth, then We1 = 0.7, We2 = 0.3), and optimize through the expansion conversion rate data later (if the expansion conversion rate of high E-value customers increases after increasing We1, then keep the adjustment, otherwise readjust), and use the formula to calculate the expansion potential indicator.

[0026] Divide the customer's annual contract amount into k levels (L1, L2,..., Lk). The level thresholds are set based on 1 / k of the average contract amount of enterprise customers (e.g., if the average amount is M, then L1 < M / k, M / k ≤ L2 ≤ 2M / k). Preset corresponding level benchmark scores for each level. The setting of the level benchmark scores is based on the creative correlation between the actual operation and maintenance value weights corresponding to the level contract amount and the return on investment ratio of enterprise resources. Statistically analyze the proportion of the average annual operation and maintenance resources (server hours, artificial service hours) consumed by customers at each level in historical data to the total operation and maintenance resources of the enterprise. The higher the proportion, the higher the level benchmark score, and the minimum value of the level benchmark score is greater than 0. For example, when k = 3, if the proportion of operation and maintenance resources consumed by customers at the L3 level (highest amount) reaches 40%, L2 reaches 35%, and L1 reaches 25%, then the benchmark scores are divided according to the ratio of 40:35:2, and the specific values are set according to historical data and actual requirements.

[0027] Through training with historical operation and maintenance resource consumption data, determine the operation and maintenance resource consumption coefficient K1 for different types of customers. The operation and maintenance resource consumption coefficient is positively correlated with the actual server hours and the number of artificial service times occupied by the customer. For example, for customers in the financial and medical industries, K1 = 1.3, and for customers in traditional manufacturing, K1 = 1.0; The standardized value of the annual contract amount corresponding to customer x is obtained by the formula and Represents the hierarchical baseline score corresponding to customer x, Represents the operation and maintenance resource consumption coefficient corresponding to customer x, Represents the maximum hierarchical baseline score.

[0028] Aiming to distinguish the essential differences between customers with long-term cooperation but unstable renewal (such as those who have cooperated for 5 years but have threatened to terminate the cooperation many times in the past 2 years) and customers with short-term cooperation but strong renewal willingness (such as those who have cooperated for 2 years but have renewed in advance), a standardized value for the cooperation duration is set, and the basic duration score is non-linearly divided according to the cooperation duration t. When t ≤ a1, the basic duration score St is set to A1, where A1 > 0. When a1 < t ≤ a2, the basic duration score , when a2 < t, , A2 = A1 + , where 、 and represent the change coefficients, which are determined based on the relationship between the cooperation duration and the input-output ratio of enterprise operation and maintenance in historical data. a1 and a2 are cooperation duration thresholds, which are set in combination with historical data and actual requirements.

[0029] Based on historical data of duration T, determine the renewal behavior of enterprises in the recent T duration, including the actual number of renewals and the number of renewals that should occur for enterprises. The number of renewals that should occur refers to the total number of renewal behaviors that should theoretically occur according to the cooperation agreement between the customer and the enterprise. Denote the ratio of the actual number of renewals to the number of renewals that should occur as the renewal stability coefficient K2. Then, the standardized value of the cooperation duration corresponding to customer x is calculated by the formula .

[0030] To avoid the disconnection between basic data and implicit dynamic data, set the comprehensive value of customer basic attributes in联动 with implicit data. The specific calculation formula is: ; where, and are used to balance and the cooperation duration in the weight of customer basic attributes, which are set according to the correlation degree of both to actual operation and maintenance resource consumption and customer retention, satisfying , is the weight of the linkage correction term, which is used to adjust the correction intensity of the operation healthiness H on the comprehensive value C of basic attributes, and is determined through business scenario analysis and historical data verification mechanism.

[0031] The customer priority scenario-based calculation module is used to analyze the interactive correlation characteristics between dynamic scheduling characteristic indicators and the comprehensive value of customer basic attributes, determine the urgency correlation factor and the value potential synergy factor, and construct a non-linear mapping model to quantify the priority of enterprise customers in combination with the real-time operation and maintenance scenario of the enterprise.

[0032] Focus on the inverse synergy between the operation health H and the resource demand intensity S, and set the urgency correlation factor F1 to quantify the urgency of the customer's current operation and maintenance needs. The formula is expressed as , where is a positive integer, set according to actual needs, and 1 is taken in this embodiment is the urgency amplification factor, dynamically calibrated based on the customer's fault diffusion speed in historical data. For example, the average duration from local anomaly to full-service interruption is used to represent the fault diffusion speed. The shorter the average duration, the faster the fault diffusion The larger the value

[0033] Capture the positive resonance between the customer's basic attributes C and the expansion potential E, and set the value potential synergy factor F2 to quantify the long-term operation and maintenance value of the customer. The formula is expressed as , where represents the synergy enhancement threshold. When E exceeds θ times of C, synergy enhancement is triggered, indicating that the customer has significant growth potential, and its long-term operation and maintenance value needs to be amplified in a non-linear manner, rather than relying solely on a single indicator of basic value or expansion potential is the potential amplification factor, calibrated based on the actual cooperation expansion data (such as the growth rate of renewal amount, the probability of new business cooperation) in historical data. The higher the growth rate of renewal amount and the probability of new business cooperation The larger the value, ensure can accurately reflect the true growth potential, and both F1 and F2 are greater than 0

[0034] Based on the multi-dimensional data linkage of business goals, resource status, and customer needs, through quantitative index monitoring and dynamic triggering mechanisms, achieve the precise division of the enterprise's real-time operation and maintenance focus, and dynamically adjust according to the enterprise's real-time operation and maintenance focus and The fusion method; The enterprise operation and maintenance cycle can be clearly divided into three categories: resource supply guarantee period, business expansion period, and regular operation and maintenance period, covering all possible operation and maintenance states. Each type of scenario generates the customer priority parameter P through and The differential fusion logic; When the server load rate collected by the resource monitoring module ≥ L0 (such as 80%), the resource supply guarantee period is triggered. The core goal is to prioritize the urgent operation and maintenance needs in the state of resource tension and avoid systemic risks caused by resource overload; When the TA period (such as 30 days) after the enterprise's new business is launched or the peak period of customer renewal (such as 2 months before the contract expires) is triggered (and the server load rate < L0), the business expansion period is triggered. The core goal is to prioritize the needs of customers with high growth potential and strengthen the long-term cooperation stickiness; All periods other than the two scenarios mentioned above are referred to as the regular operation and maintenance period. The core objective is to balance current urgent needs with long-term value and achieve a balanced allocation of operation and maintenance resources. Formulas used during resource supply guarantee period This indicates a customer priority parameter. This is a safety net factor, set based on historical data showing a reasonable resource allocation for high-F2 clients during periods of resource scarcity. Used to ensure only high And tall Customers receive a slight bonus, at which point focus The core function of (urgency) (Value potential) serves only as a backup to the basic value, to avoid high-potential but non-urgent needs occupying scarce resources; Utilizing formulas during the business expansion phase This indicates a customer priority parameter. This is a filtering coefficient, set based on the proportion of urgent needs from high-potential customers in the past. To avoid overlooking low-urgency but high-potential needs, this highlights... The weight of (value potential) (Urgency) serves only as a supplementary filter for urgent needs, ensuring that the non-urgent needs of high-potential customers also receive attention; During routine maintenance, a dynamic equilibrium model balancing urgency and potential is adopted. This model integrates the two through a geometrical mean, ensuring that urgent needs are not ignored while long-term value is not sacrificed. This is achieved using the formula... This indicates a customer priority parameter. Represents the equilibrium coefficient, based on historical data. and The distribution density of the difference is set.

[0035] By using resource monitoring microservices, differentiated allocation of resources such as computing power and manpower can be achieved based on customer P-values ​​and task requirements: For tasks from high-P value customers (P≥P1), the reserved resource pool (accounting for 20% of the total resources, used only for high-priority tasks) is automatically locked to ensure that resources are not squeezed out by low-priority tasks; For tasks of customers with medium P values ​​(P2≤P<P1), shared resources are allocated according to the proportion of P values ​​(e.g., the amount of resources allocated to a customer with P=60 is 1.2 times that of a customer with P=50). For tasks of low-P-value customers (P < P2), when resources are scarce (availability < R0, such as 30%), a resource allocation mechanism is triggered to temporarily transfer resources from low-P-value tasks to high-P-value tasks. Resources will be restored when they are sufficient to avoid service interruption for high-P-value customers.

[0036] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

[0037] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to specific implementations. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A micro-service based enterprise customer operation and maintenance management system, characterized in that, Comprise: Implicit dynamic data collection module, for analyzing the number of system operation data of customers in historical time, according to the distribution characteristics of system operation number data, different types of customers are distinguished and different data collection periods are set for different types of customers, and in each collection period, implicit dynamic data affecting customer operation and maintenance priority is collected from multiple dimensions; Implicit dynamic data includes system behavior implicit data and interaction depth implicit data; System behavior implicit data includes repeated operation times N1, invalid operation proportion R1, non-peak period resource request frequency F1 and burst resource occupation duration T1; Interaction depth implicit data includes function extension consultation frequency N2 and effective browsing time T2; The collected data is denoised and standardized, and the value range of all numerical values after standardization is between [0, M], M is set according to actual demand; Implicit dynamic feature extraction module, based on the correlation of historical operation and maintenance data, set parameters to extract feature indexes of implicit dynamic data, determine the standard value of annual contract amount and the standard value of cooperation time, and link to correct the basic attribute comprehensive value; Customer priority scenario calculation module, for analyzing the interactive correlation characteristics of dynamic scheduling feature indexes and customer basic attribute comprehensive value settings, determining the emergency degree correlation factor and the value potential synergy factor, and combining with the real-time operation and maintenance scene of enterprise, constructing a nonlinear mapping model to quantify the priority of enterprise customers. 2.The micro-service-based enterprise customer operation and maintenance management system according to claim 1, wherein, The number of system operation data of customers in historical time is analyzed, different types of customers are distinguished according to the distribution characteristics of system operation number data, and different data collection periods are set for different types of customers, the specific method is: Define the basic window with T0 as the reference period, extract the system operation times of all customers in historical T time, set the high active customer threshold X equal to the average value of the daily operation times of all customers plus the standard deviation, and the low active customer threshold Y equal to the average value of the daily operation times of all customers minus the standard deviation, the customers with daily system operation times > X are recorded as high active customers, the shrink window T1 = T0 × a, a ∈ (0, 1), a is the high active window shrink factor, the setting of a depends on the real-time data requirement and system load, the customers with daily system operation times < Y are recorded as low active customers, the expansion window T2 = T0 × b, b > 1, b is the low active window expansion factor, the setting of b is based on the data update inertia of low active customers, when the daily system operation times are in the interval [Y, X], the basic window is used. 3.The micro-service-based enterprise customer operation and maintenance management system of claim 1, wherein, Based on the correlation of historical operation and maintenance data, set parameters to extract feature indexes of implicit dynamic data, the specific method is: Based on the correlation between N1, R1 and customer complaint rate in historical operation cases, parameters Wh1 and Wh2 are set, and the formula is used to calculate the operation health index. Based on the correlation degree of F1, T1 and expansion urgency in historical resource expansion data, parameters Ws1 and Ws2 are set, and the formula The resource demand intensity index is calculated. Based on the correlation between N2, T2 and the growth of the final renewal amount in the historical cooperation expansion cases, parameters We1 and We2 are set, and the formula is used to calculate the expansion potential index. 4.The micro-service-based enterprise customer operation and maintenance management system of claim 1, wherein, Determine the standard value of annual contract amount, the specific method is: Divide the customer annual contract amount into k levels, L1, L2,..., Lk, the level threshold is set based on 1 / k of the average contract amount of enterprise customers, based on the actual operation and maintenance value weight corresponding to the level contract amount and the enterprise resource input return ratio, set the corresponding level benchmark for each level, the specific value is set according to historical data and actual demand, Determine the operation and maintenance resource consumption coefficient K1 of different types of customers through historical operation and maintenance resource consumption data training; The annual contract amount standardization value corresponding to the customer x is obtained by the formula , represents the level reference score corresponding to the customer x, represents the operation and maintenance resource consumption coefficient corresponding to the customer x, represents the maximum level reference score.

5. The microservice-based enterprise customer operation and maintenance management system according to claim 1, characterized in that, Determine the cooperation year standardized value, the specific method is: The basic years score is non-linearly divided according to the cooperation years t. When t ≤ a1, the basic years score St is set to A1, where A1 > 0. When a1 < t ≤ a2, the basic years score is set. When a2 < t, , A2 = A1 + , where and and represent the change coefficients, which are determined by combining the relationship between the cooperation years and the input-output ratio of enterprise operation and maintenance in historical data. a1 and a2 are the cooperation years thresholds, which are set by combining historical data and actual requirements; Based on historical T length data, the actual number of renewals of the enterprise and the number of renewals are determined, and the ratio of the actual number of renewals to the number of renewals is recorded as the renewal stability coefficient K2, then the standardized value of the cooperation time limit corresponding to the customer x is calculated by the formula . 6.The micro-service based enterprise customer operation and maintenance management system of claim 1, wherein, Linkage correction generates basic attribute comprehensive value, the specific method is: Customer base attribute comprehensive value, wherein, With is used to balance With cooperation years In the weight of customer base attributes, according to the correlation degree of the two to the actual operation and maintenance resource consumption and customer retention, meet , The weight of linkage correction term is determined by business scenario analysis and historical data.

7. The microservice-based enterprise customer operation and maintenance management system according to claim 1, characterized in that, Determine the emergency degree correlation factor and the value potential synergy factor, the specific method is: An emergency degree correlation factor F1 is represented by a formula , which is used to quantify the emergency degree of the current operation and maintenance demand of the customer, wherein is a positive integer, which is set according to actual demand, is an emergency degree amplification coefficient, which is dynamically calibrated based on the fault diffusion speed of the customer in historical data; The value potential synergy factor F2 is represented by the formula The long-term operation and maintenance value of the customer is quantified, wherein The synergy enhancement threshold C is represented by the formula The potential amplification coefficient is greater than 0. 8.The micro-service based enterprise customer operation and maintenance management system of claim 1, wherein, Combine enterprise real-time operation and maintenance scene, build nonlinear mapping model to quantify enterprise customer priority, the specific method is: The enterprise operation period is divided into resource supply period, business expansion period and routine operation period, and the customer priority parameter P is generated through the differentiated fusion logic of each type of scene and the difference between the two. and ​ When the server load rate collected by the resource monitoring module is greater than or equal to L0, it is the resource supply period; When the server load rate is less than L0, but the enterprise new business is online or the customer renewal peak period in TA period, it is the business expansion period, and the TA is the time length set according to the actual situation; All periods except the above two types of scenes are recorded as the regular operation and maintenance period. 9.The micro-service based enterprise customer operation and maintenance management system of claim 1, wherein, Each type of scenario is fused by with the differentiated fusion logic of the difference to generate customer priority parameter P, the specific method is: In the resource supply period, the utilization formula is The customer priority parameter is represented by The bottom coefficient is based on the reasonable resource proportion of high F2 customers during resource shortage in historical data. In the business expansion period, use the formula Indicates the customer priority parameter, is the filtering coefficient, which is set based on the proportion of high-potential customers' urgent needs in history; In the routine operation period, the formula is used to express the customer priority parameter, is used to express the balancing coefficient, which is set based on the distribution density of the difference between and .

Images