Enterprise digital diagnosis system based on AI drive

Abstract

The invention provides an enterprise digital diagnosis system based on AI driving. According to the method, enterprise internal data and enterprise external data are acquired through a data acquisition module, a digital evaluation module formulates indexes and weights of all dimensions according to the enterprise internal data, scores of all the dimensions are calculated according to the indexes and the weights, and then comprehensive scores are calculated according to the scores of all the dimensions. And the recommendation strategy optimization module generates a digital transformation recommendation strategy according to the enterprise internal data, the corresponding scores of the dimensions and the enterprise external data, and optimizes the enterprise digital transformation recommendation strategy through a reinforcement learning model so as to improve the digital transformation effect of the enterprise. The enterprise digitalization is diagnosed and evaluated through the system provided by the invention, the personalized and targeted digitalized transformation recommendation strategy is generated, the enterprise digitalized transformation recommendation strategy is optimized through the reinforcement learning model, the model can continuously learn and adjust according to feedback, and the effect of the recommendation strategy is gradually improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an enterprise digital diagnosis system driven by AI. Background Art

[0002] Intelligent transformation and digital transformation diagnosis can comprehensively evaluate the current development level of enterprises, clarify the gaps and shortcomings, formulate transformation paths according to the enterprise's conditions, guide enterprises to benchmark advanced ones, and break through development bottlenecks. The diagnosis results can provide accurate basis for enterprise policy formulation, optimize resource allocation, and create a good ecology conducive to the digital transformation of manufacturing.

[0003] The diagnostic evaluation of enterprise digital transformation is a means of evaluating the results of enterprise digital transformation. In the existing technology, the diagnostic evaluation of enterprise digital transformation is usually determined by expert scoring, and the indicator score is determined by questionnaire surveys. In addition, the existing digital transformation maturity evaluation methods are mostly based on a single perspective, and all have certain limitations. How to evaluate the results of enterprise digital transformation from multiple aspects and how to more conveniently formulate new enterprise digital transformation strategies based on the evaluation and diagnosis results to improve the digital transformation effect of enterprises have become urgent problems to be solved. Summary of the invention

[0004] The present invention describes an AI-driven enterprise digital diagnosis system that can solve the above technical problems.

[0005] According to a first aspect, an AI-driven enterprise digital diagnosis system is provided, the system comprising: Data acquisition module, used to acquire internal and external enterprise data; A digital assessment module is used to formulate indicators and weights of various dimensions based on internal enterprise data, calculate scores of various dimensions based on the indicators and weights, and then calculate a comprehensive score based on the scores of various dimensions; wherein the scores of various dimensions are used to formulate rewards in a reinforcement learning model, and the comprehensive score is used to evaluate the current digital capabilities of the enterprise; The recommendation strategy optimization module is used to generate digital transformation recommendation strategies based on the internal data of the enterprise and the scores of the corresponding dimensions and the external data of the enterprise, and optimize the enterprise digital transformation recommendation strategies through the reinforcement learning model to improve the digital transformation effect of the enterprise; The reinforcement learning model includes state, agent, action and reward. The state data includes current enterprise internal data and enterprise external data. The state data is input into the reinforcement learning model. The reinforcement learning model determines the action data corresponding to the state data. The action data includes the recommendation plan selected according to the current strategy. The agent applies the action data to the enterprise to obtain the effect after the implementation of the recommendation plan and calculates the immediate reward. The reward is formulated based on the score of each dimension of the enterprise internal data. The score of each dimension corresponding to the new enterprise internal data is obtained. The agent strategy parameters are updated according to the reward and the new state. The effect of the recommendation strategy is verified through A / B testing, and the actual effects of different recommendation strategies are compared to select the optimal strategy.

[0006] In some embodiments, the internal enterprise data includes business process data, financial system data, IT infrastructure data, customer experience data, employee skills and training data, compliance and risk management data, supply chain and logistics data.

[0007] In some embodiments, the enterprise external data includes market and industry data, customer and consumer data, economic and policy data, technology development and trend data, supply chain and partner data.

[0008] In some embodiments, the system further includes a data processing module for performing data cleaning and normalization processing on internal enterprise data and external enterprise data.

[0009] In some embodiments, the system further includes a data storage module for storing internal enterprise data and external enterprise data.

[0010] The present invention provides an enterprise digital diagnosis system driven by AI. The internal data and external data of the enterprise are acquired through the data acquisition module, and the indicators and weights of each dimension are formulated according to the internal data of the enterprise through the digital evaluation module, and the scores of each dimension are calculated according to the indicators and weights, and then the comprehensive score is calculated according to the scores of each dimension. The digital transformation recommendation strategy is generated according to the internal data of the enterprise and the scores of each dimension corresponding to it and the external data of the enterprise through the recommendation strategy optimization module, and the enterprise digital transformation recommendation strategy is optimized through the reinforcement learning model to improve the digital transformation effect of the enterprise. Through the system provided by the present invention, the decision-making efficiency and transformation effect of the enterprise can be improved from multiple angles. By calculating the scores of each dimension and generating a comprehensive score, the enterprise can fully understand its current situation and advantages, and provide a basis for formulating strategies. The recommendation strategy optimization module can generate a digital transformation recommendation strategy to ensure the personalization and pertinence of the recommendation plan. The enterprise digital transformation recommendation strategy is optimized through the reinforcement learning model. The model can continuously learn and adjust according to feedback to gradually improve the effect of the recommendation strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 A schematic diagram of the process framework of an AI-driven enterprise digital diagnosis system provided in an embodiment of this specification is shown. DETAILED DESCRIPTION

[0013] The solution provided in this specification is described below in conjunction with the accompanying drawings.

[0014] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0015] In the description of the embodiments of the present application, words such as "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a concrete way.

[0016] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, B exists alone, and A and B exist at the same time. In addition, unless otherwise specified, the term "plurality" means two or more.

[0017] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprises", "has" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0018] Figure 1 A schematic diagram of the process framework of the enterprise digital diagnosis system driven by AI provided in the embodiment of this specification is shown as follows: Figure 1 As shown, the system includes: The data acquisition module S101 is used to acquire internal data and external data of the enterprise.

[0019] Internal data can be obtained directly from internal systems through API interfaces, such as CRM systems, HR systems, etc. If the company has an internal database, it can directly obtain data through SQL queries or other database access tools. External data can be obtained from public data sources, business data providers, social media and network platforms, partners and suppliers, third-party data services, etc.

[0020] In one embodiment, the internal enterprise data includes business process data, financial system data, IT infrastructure data, customer experience data, employee skills and training data, compliance and risk management data, supply chain and logistics data.

[0021] Business process data covers various processes in the daily operation of an enterprise, including production, sales, procurement, customer service, etc. Financial system data includes information such as the enterprise's financial statements, budgets, costs, revenues, profits, etc. IT infrastructure data includes operating status and performance data on servers, network equipment, applications, security systems, etc. Customer experience data includes information such as customer satisfaction, feedback, purchasing behavior, service requests, etc. Employee skills and training data includes information such as employee skill levels, training records, performance appraisals, etc. Compliance and risk management data includes the company's compliance with laws and regulations, risk assessment reports, internal control measures, etc. Supply chain and logistics data includes information such as supplier information, inventory management, transportation and distribution, etc.

[0022] In one embodiment, the enterprise external data includes market and industry data, customer and consumer data, economic and policy data, technology development and trend data, supply chain and partner data.

[0023] Market and industry data include information such as market size, growth rate, competition landscape, market share, industry trends, etc. Customer and consumer data include information such as customer purchasing behavior, preferences, feedback, social media interactions, etc. Economic and policy data include information such as macroeconomic indicators, policies and regulations, fiscal policy, monetary policy, etc. Technology development and trend data include information such as the development of new technologies, application cases, market acceptance, future trends, etc. Supply chain and partner data include information such as supplier information, logistics data, partner performance, market coverage, etc.

[0024] In one embodiment, the AI-driven enterprise digital diagnosis system also includes a data processing module for performing data cleaning and normalization on internal and external enterprise data.

[0025] First, the internal and external data of the enterprise are cleaned. The cleaning process includes missing value processing, such as checking and filling or deleting missing records in production, sales, procurement and other processes, outlier processing, such as identifying and processing abnormal production volume, sales volume, procurement volume and other data, duplicate data processing, such as deleting duplicate orders, production batches and other records, data type conversion, such as ensuring that the data types of fields such as date, quantity, and amount are correct. Then normalization processing is performed, such as using the minimum-maximum normalization method to scale the internal and external data of the enterprise to the range of [0, 1] or [-1, 1].

[0026] In one embodiment, the AI-driven enterprise digital diagnosis system also includes a data storage module for storing internal enterprise data and external enterprise data.

[0027] Data can be divided into structured data and unstructured data, and relational databases (RDBMS), NoSQL databases, data warehouses, time series databases, etc. can be used to store internal and external data of the enterprise.

[0028] The digital evaluation module S102 is used to formulate indicators and weights of each dimension based on the internal data of the enterprise, calculate the score of each dimension based on the indicators and weights, and then calculate the comprehensive score based on the score of each dimension. Specifically, the score of each dimension is used to formulate the reward in the reinforcement learning model, and the comprehensive score is used to evaluate the current digitalization capability of the enterprise.

[0029] First, define specific indicators for each dimension. These indicators should be able to reflect the key performance indicators of the dimension. For example: Business process data includes: degree of automation, process efficiency, error rate, etc. The defined indicators are as follows: Degree of automation: the proportion of the process that is automated.

[0030] Process efficiency: the time it takes for a process to be completed.

[0031] Error rate: How often errors occur in a process.

[0032] Financial system data includes: cost control, revenue growth, profit margin, etc. The defined indicators are as follows: Cost control: the deviation between actual cost and budget.

[0033] Revenue Growth: Annual revenue growth rate.

[0034] Profit margin: Net profit as a percentage of total revenue.

[0035] IT infrastructure data includes: system availability, response time, security incidents, etc. The defined indicators are as follows: System availability: The percentage of uptime of the system.

[0036] Response time: The average time it takes the system to respond to requests.

[0037] Security incidents: The number and severity of security incidents.

[0038] Customer experience data includes: customer satisfaction, complaint rate, repeat purchase rate, etc. The defined indicators are as follows: Customer satisfaction: Customer satisfaction ratings obtained through questionnaires.

[0039] Complaint rate: number of complaints per thousand customers.

[0040] Repeat purchase rate: The proportion of customers who make repeat purchases.

[0041] Employee skills and training data include: training coverage, skills improvement, employee satisfaction, etc. The defined indicators are as follows: Training coverage: the percentage of employees who receive training.

[0042] Skill improvement: The degree to which employees’ skills are improved.

[0043] Employee Satisfaction: Employees rate their satisfaction with training and development.

[0044] Compliance and risk management data include: compliance rate, risk events, internal control, etc. The defined indicators are as follows: Compliance rate: the proportion of projects that comply with laws, regulations and company policies.

[0045] Risk events: The number and severity of risk events.

[0046] Internal Control: The effectiveness of internal control measures.

[0047] Supply chain and logistics data include: inventory turnover, on-time delivery, supplier reliability, etc. The defined indicators are as follows: Inventory turnover ratio: How often inventory turns over.

[0048] On-time delivery rate: the proportion of deliveries made on time.

[0049] Supplier reliability: The percentage of suppliers that deliver on time.

[0050] Secondly, determine the weight of each indicator, specifically assign a weight to each indicator to reflect its importance in that dimension. The sum of the weights should be 100%. For example, for the "business process" dimension: automation level: 40%, process efficiency: 30%, error rate: 30%; for the "financial system" dimension: cost control: 30%; revenue growth: 40%; profit margin: 30%. It should be understood that the weights assigned to each indicator can be formulated according to actual conditions and are not limited to those indicated in the example. The weights of other indicators will not be repeated here.

[0051] Then, collect the actual data for each indicator and calculate the score of each indicator according to the preset standard or benchmark. For example, a linear scoring method or a standardized scoring method can be used. The linear scoring method is as follows:

[0052] Among them, the actual value is X, the maximum value is M, and the minimum value is m.

[0053] Then, use the score and corresponding weight of each indicator to calculate the weighted score of each dimension. For example, for the "Business Process" dimension: automation score: 80, process efficiency score: 70, error rate score: 90. The weighted score is calculated as follows: Business process score = (80 × 0.4) + (70 × 0.3) + (90 × 0.3) = 32 + 21 + 27 = 80 Finally, the comprehensive score of the enterprise is calculated based on the weighted score of each dimension and the corresponding weight. Assume that the weight of each dimension is: business process: 20%, financial system: 20%, IT infrastructure: 15%, customer experience: 20%, employee skills and training: 10%, compliance and risk management: 10%, supply chain and logistics: 15%. The weighted scores of each dimension are: business process: 80, financial system: 75, IT infrastructure: 85, customer experience: 90, employee skills and training: 80, compliance and risk management: 95, supply chain and logistics: 80. The comprehensive score is calculated as follows: Comprehensive score = (80 × 0.2) + (75 × 0.2) + (85 × 0.15) + (90 × 0.2) + (80 × 0.1) + (95 × 0.1) + (80 × 0.15) = 81.25 The recommendation strategy optimization module S103 is used to generate a digital transformation recommendation strategy based on the internal data of the enterprise and the scores of each dimension corresponding to it and the external data of the enterprise, and optimize the enterprise digital transformation recommendation strategy through a preset reinforcement learning model to improve the digital transformation effect of the enterprise. Specifically, the reinforcement learning model includes state, agent, action and reward. The state data includes the current internal data of the enterprise and the scores of each dimension corresponding to it and the external data of the enterprise. The state data is input into the reinforcement learning model, and the reinforcement learning model determines the action data corresponding to the state data. The action data includes the recommendation scheme selected according to the current strategy. The agent applies the action data to the enterprise to obtain the effect after the implementation of the recommendation scheme and calculate the immediate reward. Specifically, the reward is formulated based on the scores of each dimension of the internal data of the enterprise, and the scores of each dimension corresponding to the new internal data of the enterprise are obtained. The agent strategy parameters are updated according to the reward and the new state, and the effect of the recommendation strategy is verified through A / B testing, and the actual effects of different recommendation strategies are compared to select the optimal strategy.

[0054] Before establishing the reinforcement learning model, it is necessary to define the reinforcement learning environment model. First, it is necessary to define the state space and define all possible states, including internal states and external states. The definition of the internal state is the same as the definition of each dimension corresponding to the internal data of the enterprise in the content of the digital evaluation module S102, and will not be repeated here. Its external state is defined according to the external data of the enterprise, for example: market and industry data are defined as: market size, competition pattern, industry trends, etc. Customer and consumer data are defined as: customer preferences, consumption habits, market feedback, etc. Economic and policy data are defined as: GDP growth rate, policy changes, tax policies, etc. Technology development and trend data are defined as: emerging technologies, technology maturity, technology application cases, etc. Supply chain and partner data are defined as: partner performance, supply chain stability, cooperation opportunities, etc. Secondly, it is necessary to define the action space and define all possible actions, such as: internal actions: optimize business processes, adjust financial strategies, upgrade IT infrastructure, improve customer experience, train employees, strengthen compliance management, optimize supply chain management, etc. External actions: adjust market positioning, optimize customer relationship management, respond to policy changes, adopt new technologies, find new partners, etc. Finally, a reward function needs to be defined to quantify the quality of the state change after each action. The reward can be based on the scores of various dimensions corresponding to its internal data, such as: business process efficiency: shortened task completion time, reduced error rate, etc.; financial performance: increased revenue, reduced costs, improved profit margins, etc.; IT infrastructure: long system uptime, short average system response time to requests, and small number of security incidents; customer experience: improved customer satisfaction, reduced complaint rate, increased repurchase rate, etc.; employee skills and training: improved employee skills, significant training results, reduced employee turnover, etc.; compliance and risk management: passed compliance checks, reduced risk events, etc.; supply chain and logistics: improved supplier performance, optimized inventory levels, reduced logistics costs, etc. For example, its reward function can be expressed as:

[0055] Among them, S is the current state, a is the current action, s′ is the new state after executing action a, S i (s′) is the score of the i-th dimension under the new state s′, w i is the weight of the i-th dimension and can be adjusted according to the needs of the enterprise.

[0056] After defining the reinforcement learning environment model, select the appropriate reinforcement learning algorithm, such as Deep Q-Networks (DQN), Policy Gradients, etc. Input the state data, i.e., the internal data of the enterprise and the external data of the enterprise, into the set reinforcement learning model. The reinforcement learning model determines the action data corresponding to the state data. The action data includes the recommended solutions selected according to the current strategy, such as: Optimize business processes: adjust the workflow to reduce the time to complete tasks; introduce automation tools to improve process efficiency. Adjust financial strategies: optimize budget allocation to increase profit margins. Increase cash flow management to reduce capital occupation. The recommended solutions can be varied, depending on the goals and current status of the enterprise, which will not be repeated here. The intelligent agent applies the action data to the enterprise to obtain the effect of the recommended solution after implementation, such as: business process data: new production line efficiency, task completion time, process efficiency, etc. Calculate the immediate reward according to the reward function, obtain the scores of each dimension corresponding to the new internal data of the enterprise, update the intelligent agent strategy parameters according to the reward and the new state, conduct A / B testing in a small range, compare the actual effects of different recommended strategies, and select the optimal strategy. Specifically, A / B testing (also known as split testing or controlled experiments) is a common method used to compare the effects of two or more versions (version A and version B) to determine which version performs better. In the context of enterprise digital transformation, A / B testing can help enterprises verify the actual effects of different recommendation strategies. For the Deep Q-Networks (DQN) algorithm, the following update agent strategy rules can be used:

[0057] in, is the Q value of the current state-action pair, α is the learning rate, γ is the discount factor, is the maximum Q value in the new state.

[0058] The reinforcement learning model provided by this solution can generate personalized recommendation strategies based on the company's internal and external data, ensuring that each strategy is tailored to the company's current specific situation. The strategy can be dynamically adjusted according to real-time data and environmental changes to ensure the timeliness and effectiveness of decision-making. In addition, the reinforcement learning model can gradually optimize the recommendation strategy through continuous learning and feedback to ensure continuous improvement of the strategy. Through the reinforcement learning model, enterprises can quickly generate and adjust recommendation strategies and improve their response speed to market changes.

[0059] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, firmware or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0060] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. An AI-driven enterprise digital diagnosis system, characterized by: The system comprises: Data acquisition module, used to acquire internal and external enterprise data; A digital assessment module is used to formulate indicators and weights of various dimensions based on internal enterprise data, calculate scores of various dimensions based on the indicators and weights, and then calculate a comprehensive score based on the scores of various dimensions; wherein the scores of various dimensions are used to formulate rewards in a reinforcement learning model, and the comprehensive score is used to evaluate the current digital capabilities of the enterprise; The recommendation strategy optimization module is used to generate digital transformation recommendation strategies based on the internal data of the enterprise and the scores of the corresponding dimensions and the external data of the enterprise, and optimize the enterprise digital transformation recommendation strategies through the reinforcement learning model to improve the digital transformation effect of the enterprise; The reinforcement learning model includes state, agent, action and reward, wherein the state includes state data, wherein the state is defined according to current enterprise internal data and enterprise external data, wherein the state data includes current enterprise internal data and enterprise external data, and the state data is input into the reinforcement learning model, wherein the reinforcement learning model determines the action data corresponding to the state data, wherein the action data includes the recommendation scheme selected according to the current strategy, and the agent applies the action data to the enterprise to obtain the effect after the implementation of the recommendation scheme, and calculates the immediate reward, wherein the reward is formulated according to the score of each dimension of the enterprise internal data, obtains the score of each dimension corresponding to the new enterprise internal data, updates the agent strategy parameters according to the reward and the new state, verifies the effect of the recommendation strategy through A / B testing, compares the actual effects of different recommendation strategies, and selects the optimal strategy.

2. The system according to claim 1, characterized in that The internal enterprise data includes business process data, financial system data, IT infrastructure data, customer experience data, employee skills and training data, compliance and risk management data, and supply chain and logistics data.

3. The system according to claim 1, characterized in that The external enterprise data includes market and industry data, customer and consumer data, economic and policy data, technology development and trend data, supply chain and partner data.

4. The system according to claim 1, characterized in that The system also includes a data processing module for performing data cleaning and normalization processing on internal data and external data of the enterprise.

5. The system according to claim 1, characterized in that The system also includes a data storage module for storing enterprise internal data and enterprise external data.

Images