Method and apparatus for providing an artificial intelligence chatbot supporting enhanced asset management functions

Abstract

An IT infrastructure management method that provides improved asset management functions through an artificial intelligence-based chatbot according to one embodiment may include the steps of receiving an asset management inquiry from a user terminal, analyzing the type of the received inquiry and referring to at least one of an asset history database or a maintenance history database, generating a response guide for the inquiry using an internal artificial intelligence tool based on the reference result, and providing the generated response guide through a conversational chatbot.

Description

Technology Field

[0001] The present invention relates to a method and apparatus for providing an artificial intelligence chatbot that supports improved asset management functions. Background Technology

[0002] With the recent rapid advancements in corporate digital transformation and artificial intelligence technology, asset management and maintenance systems have advanced beyond simple asset inventory management to a stage requiring AI-based decision support and automated work guidance.

[0003] Existing asset management systems performed functions such as asset registration, maintenance history recording, and partner management individually, requiring operators or personnel to directly access the system to manually input or search for data. This approach led to problems such as increased system complexity, higher management costs, and decreased operational efficiency. In particular, inquiries regarding equipment status checks, maintenance procedures, and responses to anomalies occurred repeatedly, requiring personnel to consult manuals or rely on skilled engineers every time. This resulted in slow response times and the inefficient consumption of human resources.

[0004] Meanwhile, technology that automatically presents complex maintenance procedures or configuration guides using artificial intelligence tools, including Large Language Models (LLMs), has recently been gaining attention. However, most systems are unable to naturally provide these AI functions to users in the form of chatbots and have limitations, being limited to simple search-based FAQ-level responses.

[0005] Accordingly, there is a growing need for a method to provide an appropriate AI tool-based conversational chatbot interface in response to user inquiries, in addition to existing asset management and maintenance functions. The problem to be solved

[0006] The problem that the present invention aims to solve is to provide a method and apparatus for providing an artificial intelligence chatbot that supports improved asset management functions.

[0007] However, the problems that the present invention aims to solve are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art to which the present invention pertains from the description below. means of solving the problem

[0008] An IT infrastructure management method that provides improved asset management functions through an artificial intelligence-based chatbot in one embodiment may include: receiving an inquiry related to asset management from a user terminal; analyzing the type of the received inquiry and referring to at least one of an asset history database or a maintenance history database; generating a response guide for the inquiry using an internal artificial intelligence tool based on the reference result; and providing the generated response guide through a conversational chatbot.

[0009] An IT infrastructure management device according to one embodiment includes a memory in which an IT infrastructure management program based on an artificial intelligence-based chatbot is stored; and a processor that controls the memory. The processor executes the IT infrastructure management program to receive an inquiry related to asset management from a user terminal, analyzes the type of the received inquiry, refers to at least one of an asset history database or a maintenance history database, generates a response guide for the inquiry using an internal artificial intelligence tool based on the reference result, and provides the generated response guide through a conversational chatbot, thereby providing an improved asset management function through an artificial intelligence-based chatbot. Effects of the invention

[0010] According to an embodiment of the present invention, by providing management of IT infrastructure through an artificial intelligence-based chatbot regardless of the type of equipment and manufacturer, multiple IT infrastructures can be managed integrally through a single platform.

[0011] In addition, according to an embodiment of the present invention, by providing a response to a person in charge's question based on a chatbot using an artificial intelligence tool, the efficiency of cost and time for customer support can be increased.

[0012] In addition, according to an embodiment of the present invention, by generating a maintenance report using a service level agreement, follow-up measures can be smoothly carried out in the event of a failure in the IT infrastructure. Brief explanation of the drawing

[0013] FIG. 1 is a block diagram illustrating an IT infrastructure management device according to one embodiment. FIG. 2 is a block diagram conceptually illustrating the functions of an IT infrastructure management program according to one embodiment. FIGS. 3 and 4 are examples of tools that can register maintenance details for IT infrastructure provided to a person in charge by a maintenance history management department according to one embodiment. FIGS. 5 and 6 are examples of tools that allow a person in charge to check maintenance details regarding IT infrastructure provided by a maintenance history management department according to one embodiment. FIGS. 7 and 8 are examples of tools that can manage the registration, use, and update of IT infrastructure provided by an asset management department to a person in charge according to one embodiment. FIGS. 9 to 12 are examples of a process in which a customer support department, according to one embodiment, provides a response to a question from a person in charge to the person in charge. FIG. 13 is a flowchart illustrating how an IT infrastructure management device according to one embodiment manages IT infrastructure. FIG. 14 is a flowchart illustrating a method in which a customer support department provides a response to a question to a person in charge according to one embodiment. FIG. 15 is a flowchart illustrating how an IT infrastructure management device according to one embodiment manages IT infrastructure. FIG. 16 is a flowchart illustrating a method in which a customer support department provides a response to a question to a person in charge according to one embodiment. FIG. 17 is a diagram illustrating the relationship between an artificial intelligence tool learning device and an IT infrastructure management device according to one embodiment. FIG. 18 is a diagram illustrating operations performed in a neural network (artificial intelligence tool) according to one embodiment. FIG. 19 is a diagram illustrating an example of an IT infrastructure management method that provides improved asset management functions through an artificial intelligence-based chatbot according to one embodiment. Specific details for implementing the invention

[0014] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but can be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims.

[0015] In describing the embodiments of the present invention, specific descriptions of known functions or configurations will be omitted if it is determined that such detailed descriptions could unnecessarily obscure the essence of the invention. Furthermore, the terms described below are defined in consideration of their functions in the embodiments of the present invention, and these definitions may vary depending on the intentions or practices of the user or operator. Therefore, such definitions should be based on the content throughout this specification.

[0016] The terms 'unit', 'unit', etc. used below refer to a unit that processes at least one function or operation, and this may be implemented in hardware or software, or a combination of hardware and software.

[0017] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.

[0018] FIG. 1 is a block diagram illustrating an IT infrastructure management device according to one embodiment.

[0019] Referring to FIG. 1, an artificial intelligence chatbot-based IT infrastructure management device (100) is designed to enable an IT infrastructure manager (or administrator, hereinafter collectively referred to as the manager) to manage registered IT infrastructure in an integrated manner regardless of type or manufacturer, and may include a processor (110), a transceiver (120), and a memory (130).

[0020] In this specification, IT infrastructure refers to all physical and logical components, such as hardware, software, networks, and data centers, required to build and run software applications, and may include network equipment, security equipment, servers, storage, etc.

[0021] The processor (110) can control the overall operation of the AI ​​chatbot-based IT infrastructure management device (100).

[0022] The processor (110) can receive a question from a person in charge of the IT infrastructure (i.e., a terminal controlled by the person in charge, hereinafter collectively referred to as the person in charge for convenience) using a transceiver (120), and transmit a guide in response to the question.

[0023] For convenience of explanation, the present specification describes that an AI chatbot-based IT infrastructure management device (100) receives a technical support request and transmits a guide using a transceiver (120), but is not limited thereto. That is, according to an embodiment, if the AI ​​chatbot-based IT infrastructure management device (100) includes an input / output device (not shown), the AI ​​chatbot-based IT infrastructure management device (100) may receive a question from a person in charge of IT infrastructure using the input / output device (not shown) or output a guide as a response to the question.

[0024] The memory (130) may store information necessary for the execution of the AI ​​chatbot-based IT infrastructure management program (200) and the AI ​​chatbot-based IT infrastructure management program (200).

[0025] In the present specification, an artificial intelligence chatbot-based IT infrastructure management program (200) may refer to software that includes commands programmed to provide maintenance history management, asset management, partner management, and customer support for IT infrastructure to a person in charge of IT infrastructure, so that the person in charge of IT infrastructure can manage registered IT infrastructure in an integrated manner regardless of type or manufacturer.

[0026] The processor (110) can load the AI ​​chatbot-based IT infrastructure management program (200) and information necessary for the execution of the AI ​​chatbot-based IT infrastructure management program (200) from memory (130) in order to execute the AI ​​chatbot-based IT infrastructure management program (200).

[0027] The processor (110) can execute an AI chatbot-based IT infrastructure management program (200) to provide maintenance history management, asset management, partner management, and customer support for the IT infrastructure to the person in charge.

[0028] The functions and / or operations of the AI ​​chatbot-based IT infrastructure management program (200) will be examined in detail through FIG. 2.

[0029] According to an embodiment, an AI chatbot-based IT infrastructure management device (100) can provide IT infrastructure personnel with maintenance history management, asset management, partner management, and customer support for IT infrastructure through a web-based SaaS platform. Therefore, compared to installed / closed platforms, personnel can manage maintenance history, etc. for IT infrastructure from anywhere.

[0030] FIG. 2 is a block diagram conceptually illustrating the functions of an IT infrastructure management program according to one embodiment.

[0031] Referring to FIGS. 1 and 2, the AI ​​chatbot-based IT infrastructure management program (200) may include a maintenance history management department (210), an asset management department (220), a partner management department (230), and a customer support department (240).

[0032] The maintenance history management unit (210), asset management unit (220), partner management unit (230), and customer support unit (240) illustrated in FIG. 2 are conceptual divisions of the functions of the AI ​​chatbot-based IT infrastructure management program (200) to easily explain the functions of the AI ​​chatbot-based IT infrastructure management program (200), but are not limited thereto. According to embodiments, the functions of the maintenance history management unit (210), asset management unit (220), partner management unit (230), and customer support unit (240) can be merged or separated and can be implemented as a series of commands included in a single program.

[0033] The maintenance history management unit (210) can manage the maintenance history of the IT infrastructure that has already been registered.

[0034] More specifically, the maintenance history management department (210) can provide the person in charge with a tool to register details regarding the maintenance of the IT infrastructure.

[0035] For example, the maintenance history management department (210) can provide a tool to the person in charge to input at least one of the following as details regarding the maintenance of the IT infrastructure: work type, work schedule, work name, work purpose and work content.

[0036] In addition, the maintenance history management department (210) can provide the person in charge with a tool to check the maintenance history of the IT infrastructure.

[0037] For example, the maintenance history management department (210) provides a tool to the person in charge to search for IT infrastructure, and the person in charge can easily determine when and what maintenance work was performed on the IT infrastructure by searching for the IT infrastructure to check maintenance details.

[0038] Additionally, the maintenance history management unit (210) can generate a report on the maintenance history of the IT infrastructure. The maintenance history management unit (210) can provide the generated report to the person in charge.

[0039] According to an embodiment, the maintenance history management unit (210) can generate a report using a Service Level Agreement (SLA) for IT infrastructure. For example, if an SLA for IT infrastructure is entered, the maintenance history management unit (210) can generate a report using the entered SLA. Alternatively, for example, if an SLA for IT infrastructure is not entered, the maintenance history management unit (210) can generate a report based on the searched SLA when an SLA for IT infrastructure is searched through a predetermined artificial intelligence tool (e.g., an LLM-based artificial intelligence tool) or a predetermined search tool.

[0040] For example, if performance indicators such as uptime, response time, failure occurrence, and recovery time for IT infrastructure are defined in the SLA, the maintenance history management unit (210) can generate a report by comparing the defined performance indicators with the current status of the IT infrastructure.

[0041] The asset management department (220) can manage the registration, use, and updates of IT infrastructure.

[0042] More specifically, the asset management department (220) can provide the person in charge with tools to manage the registration, use, and updates of IT infrastructure.

[0043] For example, the asset management department (220) can identify the end of life (EOL) and end of support (EOS) of the IT infrastructure and provide notifications regarding the EOL and / or EOS to the person in charge.

[0044] In addition, for example, the asset management department (220) can provide information about the update to the person in charge if there is an update to the IT infrastructure (e.g., firmware update).

[0045] The partner management department (230) can manage partners for IT infrastructure.

[0046] More specifically, the partner management department (230) can provide the person in charge with information regarding the partner company's contract, information regarding the person in charge within the partner company, and information regarding work related to the partner company. Through this, the person in charge can manage all information related to the partner company in an integrated manner.

[0047] When the customer support department (240) receives a question from the person in charge, it can generate a response to the question from the person in charge and send it to the person in charge.

[0048] More specifically, when the customer support department (240) receives a question from a person in charge, it can generate a response to the person in charge's question using a specific artificial intelligence tool (e.g., an LLM-based artificial intelligence tool) and send the generated response to the person in charge.

[0049] For example, if the question received from the person in charge is a method for dealing with a failure situation, the customer support department (240) can generate a guide for dealing with the failure situation as a response to the question and send it to the person in charge.

[0050] According to an embodiment, the artificial intelligence tool may be a model that has been fine-tuned by additionally inputting IT infrastructure-related data (e.g., manuals, SLAs, website data, etc.) into a previously trained LLM-based artificial intelligence model.

[0051] Additionally, according to an embodiment, if the customer support department (240) is unable to generate a response to a question from a person in charge using the first artificial intelligence tool, it may generate a response to a question from a person in charge using the second external artificial intelligence tool. According to an embodiment, the first artificial intelligence tool may be an internal artificial intelligence tool, and the second artificial intelligence tool may be an external artificial intelligence tool.

[0052] At this time, the inability to generate a response to a question may include not only cases where the AI ​​tool generates a response indicating that it is unanswerable, but also cases where the response generated by the AI ​​tool is not an appropriate response to the question from the person in charge.

[0053] To this end, the first artificial intelligence tool may generate an accuracy of the response while generating the response. At this time, if the accuracy is below a predetermined standard accuracy, the customer support department (240) determines that the response generated by the first artificial intelligence tool is not an appropriate response to the question of the person in charge and may regenerate the response to the question of the person in charge using the second artificial intelligence tool.

[0054] Meanwhile, if the customer support department (240) generates a response to a question from a person in charge using the second artificial intelligence tool, it can update the first artificial intelligence tool using the question from the person in charge and the response generated by the second artificial intelligence tool. For example, the customer support department (240) can fine-tune the first artificial intelligence tool by inputting the question from the person in charge and the response generated by the second artificial intelligence tool into the first artificial intelligence tool.

[0055] According to an additional embodiment, the customer support unit (240) may implement the operation of the first artificial intelligence tool by searching for and providing an answer to the response in a database in which an answer corresponding to the response is stored in advance. For example, when the customer support unit (240) receives an inquiry regarding a user's device, it may first search for an answer to the inquiry in the database, and if an answer corresponding to the inquiry exists, it may provide that answer to the user. If the customer support unit (240) does not have an answer corresponding to the inquiry entered by the user in the database, it may enter the inquiry into the second artificial intelligence tool and provide an answer corresponding to the inquiry. A person skilled in the art will understand that the second artificial intelligence tool described above may be pre-configured to output a statement that it cannot provide an answer when an inquiry unrelated to IT infrastructure management is entered.

[0056] FIGS. 3 and 4 are examples of tools that can register maintenance details for IT infrastructure provided to a person in charge by a maintenance history management department according to one embodiment.

[0057] Referring to FIGS. 1 to 4, the maintenance history management unit (210) can provide a person in charge with a registration button that can register details regarding the maintenance of the IT infrastructure.

[0058] When a person in charge selects the registration button, the maintenance history management unit (210) can provide the person in charge with an interface to enter details regarding the maintenance of the IT infrastructure. The person in charge can enter details regarding the maintenance of the IT infrastructure by selecting a work type, specifying a work schedule, and entering a work name, work purpose, and work content.

[0059] FIGS. 5 and 6 are examples of tools that allow a person in charge to check maintenance details regarding IT infrastructure provided by a maintenance history management department according to one embodiment.

[0060] Referring to FIGS. 1 to 6, when a person in charge selects an IT infrastructure that has performed maintenance or selects the maintenance details of the IT infrastructure, the maintenance history management unit (210) can enable the person in charge to check the details of the maintenance of the IT infrastructure.

[0061] At this time, if it is necessary to delete the maintenance history or modify the details, the person in charge can delete the maintenance history or modify the details of the maintenance using the interface provided by the maintenance history management department (210).

[0062] FIGS. 7 and 8 are examples of tools that can manage the registration, use, and update of IT infrastructure provided by an asset management department to a person in charge according to one embodiment.

[0063] Referring to FIGS. 1, FIGS. 2, FIGS. 7 and FIGS. 8, the asset management department (220) can provide a list of registered IT infrastructure to the person in charge.

[0064] The asset management department (220) can enable a person in charge to search for specific IT infrastructure by entering an asset number, asset name, model name, serial number, etc. Additionally, the asset management department (220) can enable a person in charge to search for IT infrastructure that meets the corresponding conditions based on equipment type, manufacturer name, product type, product family, equipment status (e.g., in use, unused, discarded, etc.).

[0065] The asset management department (220) can extract the search results into a specified file format (e.g., Excel file) and provide it to the person in charge.

[0066] When a person in charge selects an IT infrastructure for which they wish to check details, the asset management department (220) can provide the person in charge with details about the selected IT infrastructure.

[0067] At this time, the asset management department (220) can allow the person in charge to input, modify, or delete detailed information such as asset number, asset name, installation location, memo, and equipment status.

[0068] FIGS. 9 to 12 are examples of a process in which a customer support department, according to one embodiment, provides a response to a question from a person in charge to the person in charge.

[0069] Referring to FIGS. 1, FIGS. 2 and FIGS. 9 through 12, the customer support department (240) can provide a question button to the person in charge so that the person in charge can enter a question.

[0070] When a person in charge selects the question button, the customer support department (240) can provide the person in charge with an interface to input information about the IT infrastructure and the content of the question that the person in charge wishes to ask.

[0071] For example, a person in charge can input a question including equipment type, manufacturer name, product type, product family, product (detailed model), inquiry content, and inquiry details using an interface provided by the customer support department (240). For example, the person in charge can input a question in the order of equipment type, manufacturer name, product type, product family, product (detailed model), inquiry content, and inquiry details. At this time, the person in charge can input the equipment type, etc., only for IT that has already been registered.

[0072] When a person in charge inputs a question including ‘network equipment’ as the equipment type, ‘Cisco Systems’ as the manufacturer name, ‘switch’ as the product type, ‘Catalyst 9500’ as the product family, and ‘what should I do if the switch power is not turning on?’ as the details of the inquiry, the Customer Support Department (240) can use a specific artificial intelligence tool to provide the person in charge with a guide on ‘how to respond when the switch power is not turning on’ as a response to the question.

[0073] At this time, if the person in charge wishes to ask additional questions about the guide provided by the customer support department (240), they can enter additional questions through the interface provided by the customer support department (240).

[0074] If the person in charge asks an additional question such as "What is a surge protector?", the customer support department (240) can provide additional guidance to the person in charge regarding the additional question using a specific artificial intelligence tool.

[0075] FIGS. 13 and 14 are examples of a method for reporting failures and requesting technical support for IT infrastructure by entering a QR code.

[0076] Referring to FIGS. 1, FIGS. 2 and FIGS. 13, a QR code may be attached to the IT infrastructure managed by the AI ​​chatbot-based IT infrastructure management device (100).

[0077] When a failure occurs in the IT infrastructure, if a person in charge uses a smartphone to scan a QR code attached to the IT infrastructure, the AI ​​chatbot-based IT infrastructure management device (100) can provide the person in charge with a page to report the failure in the IT infrastructure. Through the page, the person in charge can easily report that a failure has occurred in the IT infrastructure.

[0078] Additionally, referring to FIG. 14, when a person in charge scans a QR code attached to the IT infrastructure using a smartphone, the AI ​​chatbot-based IT infrastructure management device (100) can provide the person in charge with a page where they can request technical support (i.e., a response to a question) for the IT infrastructure. Through this page, the person in charge can easily receive technical support for the IT infrastructure.

[0079] FIG. 15 is a flowchart illustrating how an IT infrastructure management device according to one embodiment manages IT infrastructure.

[0080] Referring to FIGS. 1, 2 and 15, an AI chatbot-based IT infrastructure management device (100) can register multiple IT infrastructures, each having a different type of equipment and manufacturer (S1500).

[0081] An AI chatbot-based IT infrastructure management device (100) can provide at least one of maintenance history management, asset management, partner management, and customer support for each of multiple IT infrastructures (S1510).

[0082] FIG. 16 is a flowchart illustrating a method in which a customer support department provides a response to a question to a person in charge according to one embodiment.

[0083] Referring to FIGS. 1, FIGS. 2 and FIGS. 16, the customer support department (240) can receive a question from a person in charge (S1600).

[0084] If the internal AI tool fails to generate a response to the question (S1610 'No'), the customer support department (240) can use an external AI tool to generate a response to the question and provide the generated response to the person in charge (S1620).

[0085] Afterwards, the customer support department (240) can update the internal AI tool using the questions from the person in charge and the responses generated by the external AI tool (S1630).

[0086] On the other hand, if an internal AI tool generates a response to a question (S1610 'Yes'), the customer support department (240) can provide the response generated using the internal AI tool to the person in charge (S1640).

[0087] The method of providing IT device management functions using the AI ​​tool described above may also be provided using an AI chatbot. More specifically, an AI chatbot-based IT infrastructure management device (100) may provide a chatbot interface to receive a user's question and generate and provide an AI-based answer to the received question. An example of providing an answer using an AI chatbot is further presented below through FIG. 19.

[0088] FIG. 17 is a diagram illustrating the relationship between an artificial intelligence tool learning device and an IT infrastructure management device according to one embodiment.

[0089] The artificial intelligence tool learning device (1700) may correspond to any computing device that performs the operation of learning the artificial intelligence tool described above.

[0090] Referring to FIG. 17, an artificial intelligence tool learning device (1700) according to one embodiment corresponds to a computing device having various processing functions such as generating a neural network, training (or learning) a neural network, or retraining a neural network. For example, the artificial intelligence tool learning device (1700) can be implemented as various types of devices such as a PC (personal computer), a server device, or a mobile device.

[0091] An artificial intelligence tool learning device (1700) can generate a trained neural network (1710) (the artificial intelligence tool described above) by repeatedly training (learning) a given initial neural network. Generating a trained neural network (1710) may mean determining neural network parameters. Here, the parameters may include various types of data input to / output to the neural network, such as input / output activations, weights, biases, etc. As the repeated training of the neural network progresses, the parameters of the neural network may be tuned to compute a more accurate output for a given input.

[0092] The artificial intelligence tool learning device (1700) can transmit the trained neural network (1710) to the IT infrastructure management device (1750). The IT infrastructure management device (1750) may be included in a mobile device, an embedded device, etc. The IT infrastructure management device (1750) may be dedicated hardware for operating the neural network.

[0093] The IT infrastructure management device (1750) can run the trained neural network (1710) as is, or run the neural network (1760) in which the trained neural network (1710) has been processed (e.g., quantized). The IT infrastructure management device (1750) running the processed neural network (1760) can be implemented in a separate, independent device from the artificial intelligence tool learning device (1700). However, it is not limited to this, and the IT infrastructure management device (1750) can also be implemented within the same device as the artificial intelligence tool learning device (1700).

[0094] FIG. 18 is a diagram illustrating operations performed in a neural network (artificial intelligence tool) according to one embodiment.

[0095] Referring to FIG. 18, a neural network (artificial intelligence tool) (1800) has a structure including an input layer, hidden layers and an output layer, and can perform operations based on received input data (e.g., I1 and I2) and generate output data (e.g., O1 and O2) based on the results of the operation.

[0096] The neural network (artificial intelligence tool) (1800) may be a DNN or an n-layer neural network including two or more hidden layers. For example, as illustrated in FIG. 18, the neural network (artificial intelligence tool) (1800) may be a DNN including an input layer (Layer 1), two hidden layers (Layer 2 and Layer 3), and an output layer (Layer 4). The DNN may include, but is not limited to, Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Deep Belief Networks, Restricted Boltzmann Machines, etc.

[0097] When the neural network (artificial intelligence tool) (1800) is implemented as a DNN architecture, it includes more layers capable of processing valid information, so the neural network (artificial intelligence tool) (1800) can process more complex data sets than a neural network with a single layer. Meanwhile, although the neural network (artificial intelligence tool) (1800) is depicted as including four layers, this is merely an example, and the neural network (artificial intelligence tool) (1800) may include fewer or more layers, or fewer or more channels. That is, the neural network (artificial intelligence tool) (1800) may include layers of various structures different from those depicted in FIG. 18.

[0098] Each layer included in the neural network (artificial intelligence tool) (1800) may include multiple channels. A channel may correspond to multiple artificial nodes known as a neuron, a processing element (PE), a unit, or similar terms. For example, as shown in FIG. 3, Layer 1 may include two channels (nodes), and Layer 2 and Layer 3 may each include three channels. However, this is merely an example, and each layer included in the neural network (300) may include various numbers of channels (nodes).

[0099] The channels included in each layer of the neural network (artificial intelligence tool) (1800) can be connected to each other to process data. For example, one channel can receive data from other channels and perform calculations, and can output the results of the calculations to other channels.

[0100] The input and output of each channel can be referred to as input activation and output activation, respectively. That is, activation can be the output of a channel and, at the same time, a parameter corresponding to the input of the channels included in the next layer. Meanwhile, each channel can determine its own activation based on the activations, weights, and biases received from the channels included in the previous layer. Weights are parameters used to calculate the output activation at each channel and can be values ​​assigned to the connection relationships between channels.

[0101] Each of the channels can be processed by a computational unit or processing element that receives an input and outputs an output activation, and the input and output of each channel can be mapped. For example, σ is an activation function, and is the weight from the k-th node in the (i-1)-th layer to the j-th node in the i-th layer, and is the bias value of the j-th node included in the i-th layer, and Let be the activation of the j-th node of the i-th layer. It can follow the following mathematical formula 1.

[0102]

[0103] As shown in FIG. 18, the activation of the first channel (CH 1) of the second layer (Layer 2) is It can be expressed as. Also, is according to mathematical formula 1 It may have a value. However, the mathematical formula 1 described above is merely an example to explain the activation, weight, and bias used to process data in a neural network (artificial intelligence tool) (1800), and is not limited thereto. The activation may be a value obtained by passing the weighted sum of activations received from the previous layer through an activation function such as a sigmoid function or a Rectified Linear Unit (ReLU) function.

[0104] According to an additional embodiment, the customer support department (240) can perform an action to continuously improve the prediction accuracy of an internal artificial intelligence model by comparing the quality of the response (guide) generated by the artificial intelligence tool with performance indicators defined in the Service Level Agreement (SLA).

[0105] More specifically, the first artificial intelligence tool can generate and provide a guide response to a question entered by the person in charge (e.g., failure situation of specific equipment, inspection procedure, etc.).

[0106] At this time, the SLA indicators transmitted from the maintenance history management department (210) include predetermined standard values ​​for individual IT equipment, such as the allowable response time, recovery time, and availability rate of the equipment, and the customer support department (240) can verify whether these SLA standards are met when the response through the AI ​​tool's response is completed. For example, the customer support department (240) can perform the verification procedure described above by comparing the indicators calculated after the completion of the response based on the guide with the previously stored SLA indicators. For example, if the response procedure presented by the AI ​​tool actually shortens the recovery time or exceeds the SLA target, the response is judged to be a "high-quality response."

[0107] The quality evaluation performed by the customer support unit (240) according to one embodiment is the loss function following the loss function calculated based on mathematical formula 2. It can be performed through.

[0108]

[0109] is a value measured from actual SLA metrics (e.g., recovery time, response success rate) at time t, is the measured or predicted SLA performance value when applying the guide response suggested by the first AI tool, and are the time rate of change of SLA performance, respectively, Is and It can be calculated based on Equation 3 as a probability distribution divergence value representing the difference in uncertainty between them.

[0110]

[0111] is the probability distribution of SLA performance metrics (e.g., recovery time, responsiveness, availability rate, etc.) actually observed at time t, may be a probability distribution of SLA performance indicators predicted or measured at the same point in time based on the guidance of the first artificial intelligence tool at time t.

[0112] The customer support department (240) can dynamically improve not only the accuracy of the response but also the SLA compliance level by fine-tuning the parameter θ of the first artificial intelligence tool to minimize the loss function according to mathematical formula 2. Through this, according to the present embodiment, the first artificial intelligence tool can be learned to provide more appropriate countermeasures by learning and evolving on its own in a direction that improves the actual service level (SLA) indicator, going beyond simply providing sentence-type answers to questions.

[0113] According to another embodiment, the partner management unit (230) can provide a means to improve the efficiency of maintenance by supporting optimized resource allocation through a mathematical optimization model by automatically determining the assignment between a plurality of maintenance requests (Task) i and a plurality of partner companies j.

[0114] More specifically, the partner management department (230) has an optimal allocation value that minimizes the objective function according to mathematical formula 4. It can produce.

[0115]

[0116] The constraints of the above mathematical formula 4 may be the same as mathematical formula 5.

[0117]

[0118] is the estimated response time when partner j processes request i, is the cost of partner J, is the past maintenance success rate or customer satisfaction indicator (average of survey figures on satisfaction), and is the average response time and standard deviation of all partner companies, These may be weighting factors reflecting the importance of time, cost, quality, and stability defined in the SLA, respectively.

[0119] As explained earlier It can be calculated based on mathematical formula 6, based on response log data collected from the past maintenance history of partner j.

[0120]

[0121] is the average past response time of partner j, is the waiting time for partner j (waiting time calculated based on the currently existing task queue), is a distance-based transmission delay function between request location Li and partner location Lj, calculated based on the distance between the two locations and the traffic congestion between the two points, and can be a weight estimated by an AI learning model.

[0122] The Partner Company Management Department (230) class While minimizing, Assignment value to maximize By determining this, it is possible to select the optimal partner and support matching. Additionally, the last term of Equation 4 can correspond to a means that takes into account the load balancing of all partners.

[0123] According to another embodiment, an AI chatbot-based IT infrastructure management device (100) can provide a risk alert if the current operating state exceeds a predetermined threshold for similarity to past failure patterns by analyzing the operating data of a specific device (e.g., current, voltage, temperature, CPU load rate, response delay, etc.) and statistically comparing it with the operating logs of similar equipment from other companies or the same industry group stored in an external database. More specifically, the processor (110) [is] the current operating state vector of the device Average state vector of an external reference group of equipment (including measurement information regarding equipment status (e.g., temperature, voltage, etc.) as individual vectors) and covariance matrix Compared to, statistical similarity score according to It can be calculated based on mathematical formula 7.

[0124]

[0125] The processor (110) can statistically calculate the average failure rate for each similarity interval (e.g., 0.0-0.2, 0.2-0.4, ..., 0.8-1.0) based on historical data accumulated from multiple external companies. More specifically, the processor (110) can calculate a conditional failure probability function based on Equation 8. It can produce.

[0126]

[0127] is the similarity score for the similarity between the k-th case in the external dataset and the current equipment, is information on whether a failure occurred in the k-th case (1 for failure, 0 for normal), is an interval function representing a similarity interval (e.g., [0.4, 0.6]), is an indicator function that returns 1 if included in the corresponding range, is a continuous probability function interpolated by a quadratic spline for the failure rate by interval, is a smoothing coefficient, which may be a value determined for abrupt correction.

[0128] is a function that takes a similarity s as input and calculates the average failure rate of a group of similar cases, and the value may tend to increase as the similarity increases (i.e., as the condition of the device is similar to past failure cases).

[0129] The processor (110) uses the above mathematical formulas 7 and 8 to determine the failure risk at the current time. It can be calculated as in mathematical formula 9.

[0130]

[0131] Here, is a sensitivity parameter for the rate of change in similarity, indicating that the device's condition deteriorates rapidly (i.e., When this decreases rapidly, it can further amplify the risk of failure.

[0132] According to an embodiment of the present invention, by providing management of IT infrastructure regardless of equipment type and manufacturer, multiple IT infrastructures can be managed integrally through a single platform.

[0133] The AI ​​chatbot-based IT infrastructure management device (100) is produced If this exceeds a preset threshold, it is determined that a failure is imminent, a prior notification is sent to the person in charge's terminal, and a "Predictive Maintenance" schedule can be automatically generated through the maintenance history management unit (210).

[0134] In addition, according to an embodiment of the present invention, by using an artificial intelligence tool to provide a response to a question from a person in charge, the efficiency of cost and time for customer support can be increased.

[0135] In addition, according to an embodiment of the present invention, by generating a maintenance report using a service level agreement, follow-up measures can be smoothly carried out in the event of a failure in the IT infrastructure.

[0136] FIG. 19 is a diagram illustrating an example of an IT infrastructure management method that provides improved asset management functions through an artificial intelligence-based chatbot according to one embodiment.

[0137] An AI chatbot-based IT infrastructure management device (100) according to one embodiment receives an inquiry related to asset management from a user terminal in step (S1900), analyzes the type of the received inquiry in step (S1910) and refers to at least one of an asset history database or a maintenance history database, generates a response guide for the inquiry using an internal AI tool based on the reference result in step (S1920), and can provide the generated response guide through a conversational chatbot in step (S1930).

[0138] For example, in step (S1900), questions in the form of natural language, such as "What is the regular inspection cycle of equipment A?" or "What is the part scheduled for replacement?", may be input into the chatbot-based IT infrastructure management device (100). In step (S1910), the chatbot-based IT infrastructure management device (100) identifies the intent of the query using a natural language processing model (NLP), classifies whether the query belongs to a category such as maintenance, asset status, partner, cost, or SLA-related inquiry, and can generate a natural language response based on information retrieved using a first artificial intelligence tool or a second artificial intelligence tool. For example, if an internal artificial intelligence tool outputs a result below a reliability standard, an external artificial intelligence tool (e.g., ChatGPT API, etc.) may be called to generate an auxiliary response, and by synthesizing these to obtain a final response, a response guide may be generated.

[0139] In step (S1930), the AI ​​chatbot-based IT infrastructure management device (100) can generate a response guide and provide it to the user.

[0140] According to an additional embodiment, the IT infrastructure management device (100) is configured to predict the response generation load based on the input frequency and context complexity of user requests, and to minimize response latency by dynamically optimizing the token generation speed and pipeline splitting ratio of the language model based on the predicted load.

[0141] For example, the request vector corresponding to the request at input time ti If defined as such, here, It can be calculated by combining the complexity of the conversational context, the proportion of rare words, the average token length, the number of subordinate clauses, and the degree of dependency on the previous conversation, and a person of ordinary technical skill will understand that the calculation method is not limited to a specific direction, and is the length of the request, may mean the frequency of occurrence of requests. The IT infrastructure management device (100) has a response load prediction function based on mathematical formula 10. It can produce.

[0142]

[0143] is a weighting factor based on the system environment, and can mean the average request interval.

[0144] Load value If the input exceeds a preset threshold, the IT infrastructure management device (100) may suggest a direction to the user to reduce the load. For example, the IT infrastructure management device (100) may provide a message such as, "Would you like to break down the question into more specific parts and input it?" In this process, the IT infrastructure management device (100) may calculate the semantic complexity of individual elements included in the input sentence (e.g., quantifying complexity based on the ratio of rare words, average token length, number of dependent clauses, and degree of dependency with previous conversations), and visually display and provide elements that can be deleted or shortened. Additionally, instead of generating a response immediately, the IT infrastructure management device (100) may provide a means to alleviate the model computation load by notifying the user of a predetermined delay time and registering the response to be delivered after a certain period of time in a waiting queue. According to an embodiment, the IT infrastructure management device (100) may provide a response first by prioritizing elements that can be processed with a low load among the requests, and then provide a message indicating that a response for the remaining elements will be provided after a predetermined period of time, and then provide it later depending on the load situation.

[0145] In addition, the IT infrastructure management device (100) may provide a means to respond to high load situations by providing a summary response first when the load value exceeds a predetermined threshold as described above, providing a preliminary response through low load, and then providing a detailed response after a predetermined waiting time.

[0146] According to an additional embodiment, the IT infrastructure management device (100) can provide a step-by-step technical support service, and to this end, can receive a technical support request from a user terminal.

[0147] The IT infrastructure management device (100) can perform primary technical support by analyzing the complexity or type of the technical support request and providing an automated response via the web or application. The method of performing primary technical support may correspond to a series of methods described later through FIGS. 3 to 19.

[0148] If it is determined that a response is impossible through primary technical support, the IT infrastructure management device (100) can perform secondary technical support by transmitting a request to the terminal of the person in charge at the AI ​​control center. For example, cases where a response is impossible through primary technical support include situations where the technical support problem is not resolved without the intervention of a technical support person, situations not included in the manual, and situations where the AI ​​tool cannot provide an appropriate answer, but may also include situations where the problem is not resolved even if an appropriate answer is provided during primary technical support due to the user's low proficiency. An ordinary technician will understand that examples of cases where a response is impossible through primary technical support are not limited to the examples presented, and may apply to any situation requiring direct support from a person in charge. If a response is impossible through primary technical support, the IT infrastructure management device (100) can perform secondary technical support by creating a communication channel to allow the person in charge at the AI ​​control center and the user to communicate. During the secondary technical support process, the person in charge at the AI ​​control center can identify the user's problem through chat, phone calls, or remote access, and provide a solution to the problem. An AI control center may refer to a center that includes IT device management experts, and there may not be any geographical restrictions.

[0149] The IT infrastructure management device (100) can perform third technical support including on-site inspection if it is determined that on-site inspection is necessary as a result of second technical support. If an AI control center manager determines that on-site inspection of an IT device is necessary during the remote support process, they can perform third technical support including on-site inspection by dispatching a suitable technical expert based on the location information where the IT device is located.

[0150] The IT infrastructure management device (100) can determine the optimal set of technicians based on a multi-purpose optimization-based field response model that integrates cost, distance, reliability, and load rate to determine which entity among multiple technical personnel or partner companies to dispatch when on-site inspection is required during the third technical support process. More specifically, the IT infrastructure management device (100) can calculate the estimated travel distance dj, estimated arrival time tj, dispatch cost cj, past success rate (reliability) rj, and current load rate uj (current work intensity) for each technical personnel j, and calculate the objective function F(j) according to Equation 11 based on the calculation results.

[0151]

[0152] Here, and is the average distance and average cost of the entire candidate group, is the SLA standard time It is a penalty function for delays exceeding, and These are weighting factors for distance, cost, reliability, load rate, and time delay, respectively, and can be dynamically adjusted according to the operational policy of the AI ​​control center and the importance of each SLA item.

[0153] The IT infrastructure management device (100) is an optimal set of technologists according to Equation 12 in a direction that minimizes the objective function according to Equation 11. It can determine and provide it to the user.

[0154]

[0155] Here, represents a load balancing term that distributes the load across the entire technical workforce to prevent excessive work from being concentrated on specific personnel, and can be the equilibrium weighting factor.

[0156] Additionally, the IT infrastructure management device (100) has an optimal path according to Equation 13 to minimize travel time when personnel are dispatched to multiple locations. It can provide.

[0157]

[0158] Here, is the order of visits, is average driving speed, is the traffic congestion function, and ρ is the traffic sensitivity coefficient.

[0159] Additionally, the IT infrastructure management device (100) is the set of optimal technicians described above. In the process of providing this, user feedback can be matched and provided based on each technician's IT equipment repair history and individual repairs. Additionally, by providing information on the parts held by each technician, a means can be provided for the user to quickly select a technician capable of immediate repair.

[0161] Combinations of each block of the block diagram attached to this specification and each step of the flowchart may be performed by computer program instructions. Since these computer program instructions may be loaded into an encoding processor of a general-purpose computer, a computer for special purposes, or other programmable data processing equipment, the instructions executed through the encoding processor of the computer or other programmable data processing equipment create means for performing the functions described in each block of the block diagram or each step of the flowchart. Since these computer program instructions may also be stored in computer-available or computer-readable memory that can be directed toward the computer or other programmable data processing equipment to implement the function in a specific way, the instructions stored in computer-available or computer-readable memory may also produce a manufactured item containing instruction means for performing the function described in each block of the block diagram or each step of the flowchart. Since computer program instructions can be loaded onto a computer or other programmable data processing equipment, instructions that execute a computer or other programmable data processing equipment by performing a series of operation steps on the computer or other programmable data processing equipment to create a process executed by the computer can also provide steps for executing the functions described in each block of the block diagram and each step of the flowchart.

[0162] Additionally, each block or each step may represent a module, segment, or part of code containing one or more executable instructions for executing a specified logical function(s). Also, it should be noted that in some alternative embodiments, the functions mentioned in the blocks or steps may occur out of order. For example, two blocks or steps described in succession may actually be performed substantially simultaneously, or the blocks or steps may sometimes be performed in reverse order according to the corresponding function.

[0163] The above description is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art to which the present invention pertains will be able to make various modifications and variations within the scope of the essential quality of the present invention. Accordingly, the embodiments disclosed in the present invention are intended to explain, not limit, the technical concept of the present invention, and the scope of the technical concept of the present invention is not limited by such embodiments. The scope of protection of the present invention shall be interpreted by the claims below, and all technical concepts within the equivalent scope shall be interpreted as being included within the scope of rights of the present invention. Explanation of the symbols

[0165] 100: AI Chatbot-based IT Infrastructure Management Device 200: AI Chatbot-based IT Infrastructure Management Program 210: Maintenance History Management Department 220: Asset Management Department 230: Partner Company Management Department 240: Customer Support Department

Claims

Claim 1 An IT infrastructure management method that provides an improved asset management function through an AI-based chatbot, performed by an IT infrastructure management device, comprises: receiving an asset management-related inquiry from a user terminal; analyzing the type of the received inquiry and referring to at least one of an asset history database or a maintenance history database; generating a response guide for the inquiry using an internal AI tool based on the reference result; and providing the generated response guide through a conversational chatbot. The IT infrastructure management method further comprises a step of providing phased technical support. The step of providing phased technical support includes performing first-stage technical support by analyzing the complexity and type of the technical support request received from the user terminal and providing an automated response via a web or application in response to the technical support request received from the user terminal; if it is determined that a response is impossible through the first-stage technical support, transmitting the request to a terminal of a person in charge at an AI control center and performing second-stage technical support by providing technical support with the assistance of the person in charge at the AI ​​control center; and if it is determined that a response is impossible through the second-stage technical support, performing third-stage technical support by dispatching multiple technical personnel or personnel from a partner company. In the process of performing the above third-level technical support, the IT infrastructure management device calculates an optimal set of technicians based on mathematical formulas 1 and 2, determines one of the technicians included in the optimal set of technicians as the technician to be dispatched, and performs the above third-level technical support, wherein mathematical formula 1 is And, above F(j) is the objective function, above dj is the estimated travel distance, above tj is the estimated time of arrival, above cj is the dispatch cost, above rj is the past success rate, above uj (current work intensity) is the current load rate, and above and above is the average distance and average cost of the entire candidate group, and the above is the SLA standard time It is a penalty function for delays exceeding, and the above are weighting factors for distance, cost, reliability, load rate, and time delay, respectively, which are dynamically adjusted according to the operational policies of the AI ​​control center and the importance of each SLA item, and the above mathematical formula 2 is, and, above is a set of optimal technologists, and the above is a load balancing term that distributes the load across the entire technical workforce to prevent excessive work from being concentrated on specific personnel, and An IT infrastructure management method that provides improved asset management functions through an AI-based chatbot, which is a balanced weighting factor. Claim 2 delete

Images