Device for generating cloud infrastructure based on artificial intelligence and method thereof

Abstract

The present invention relates to an artificial intelligence-based cloud infrastructure creation device and method. The problem to be solved is to enable users to easily configure various cloud infrastructures and provide diverse options for adopting and utilizing the cloud, and to support safe cloud adoption by creating infrastructure based on Infrastructure as Code (IaC), which is advantageous for cloud infrastructure management, thereby preventing human errors and reducing management points. For example, an AI-based cloud infrastructure creation device is disclosed, comprising: a cloud infrastructure creation request unit that receives and uploads an image to request the creation of cloud infrastructure; an image-text conversion unit that converts the image into text using a pre-trained generative AI model; a cloud infrastructure information generation unit that extracts cloud infrastructure resources based on the text and generates correlation information of the extracted cloud infrastructure resources; and a cloud infrastructure creation unit that creates cloud infrastructure based on Infrastructure as Code (IaC) according to cloud infrastructure resources and correlation information.

Description

Technology Field

[0001] An embodiment of the present invention relates to an artificial intelligence-based cloud infrastructure creation device and method. Background Technology

[0002] With the advent of the digital era and the rapid changes in the business environment, there is an increasing need for IT infrastructure capable of corporate flexibility and agility, and the cloud is being utilized across various fields to meet these corporate needs.

[0003] In particular, as remote work and non-face-to-face services increase, cases of developing and adopting cloud solutions are on the rise, and cloud environments that provide flexible resources to support the activation of AI are emerging as an attractive option.

[0004] Cloud refers to "service provider servers" based on the convention of representing computing service provider servers as a cloud shape. It involves storing software and data on a central computer connected to the Internet, allowing users to access data anytime and anywhere simply by connecting to the Internet.

[0005] According to the Ministry of Employment and Labor, there is a projected shortage of approximately 18,800 cloud personnel by 2027. The shortage is intensifying due to the increasing demand for personnel across all areas from 'operations' to 'development,' the spread of cloud services such as servers and software, and continuous market growth. In particular, there is a shortage of both operational personnel for server management and maintenance, as well as system development personnel.

[0006] According to the IDG 2023 Survey and Forecast of Domestic Cloud Computing Adoption, technical solutions for cloud-native environments are assessed to already be in place, but specialized personnel such as cloud architects and security managers are required to actually implement these technical solutions, and detailed difficulties such as resource and time management, labor costs, and prioritization are assessed.

[0007] As such, at a time when the expansion and adoption of cloud services need to be promoted, there is a problem in that a shortage of cloud specialists and difficulties in cloud adoption are anticipated in the future.

[0008] Infrastructure as Code (IaC) is the management and provisioning of infrastructure through code rather than manual processes, and currently, the de facto industry standard services representing Infrastructure as Code (IaC) include Terraform and Ansible.

[0009] Utilizing IaC generates configuration files containing cloud infrastructure specifications, making it easy to edit and deploy configurations and ensuring that the same environment is provisioned every time. Additionally, since the infrastructure is managed in the form of code, version control is possible, which is why continued utility is expected for future cloud infrastructure deployments. Prior art literature

[0010] Published Patent Application No. 10-2024-0003316 (Publication Date: January 8, 2024) Published Patent Application No. 10-2015-0074411 (Publication Date: July 2, 2015) Registered Patent Application No. 10-2438067 (Registration Date: August 25, 2022) Registered Patent Application No. 10-1505468 (Registration Date: March 18, 2015) The problem to be solved

[0011] Embodiments of the present invention provide an AI-based cloud infrastructure generation device and method that allow for easy configuration of various cloud infrastructures and offer diverse options to users who wish to adopt and utilize the cloud, and support safe cloud adoption by generating infrastructure based on Infrastructure as Code (IaC), which is advantageous for cloud infrastructure management, thereby preventing human errors and reducing management points. means of solving the problem

[0012] An artificial intelligence-based cloud infrastructure generation device according to one embodiment of the present invention includes: a cloud infrastructure generation request unit that receives an image as input and uploads it to request the generation of cloud infrastructure; an image-text conversion unit that converts the image into text using a pre-trained generative artificial intelligence model; a cloud infrastructure information generation unit that extracts cloud infrastructure resources based on the text and generates correlation information of the extracted cloud infrastructure resources; and a cloud infrastructure generation unit that generates cloud infrastructure based on Infrastructure as Code (IaC) according to cloud infrastructure resources and correlation information.

[0013] In addition, it may further include a cloud infrastructure architecture provider that diagrams the architecture for the cloud infrastructure based on the code of the cloud infrastructure and provides the diagrammed cloud infrastructure architecture.

[0014] In addition, the cloud infrastructure creation request unit can receive an image file in a raster graphic format provided by the user.

[0015] In addition, the cloud infrastructure creation request unit can generate an image file in a vector graphic format using a drawing tool that provides physical and virtual resource image templates representing cloud infrastructure resources, and block and line shape templates for diagramming the correlation between each image.

[0016] In addition, the image-to-text conversion unit can recognize physical and virtual resource images representing cloud infrastructure resources, and block and line shapes representing the correlation between each resource, respectively, and convert the recognized information into text.

[0017] Additionally, it may further include a cloud infrastructure detailed setting request unit that receives a text-based prompt from a user based on a pre-trained language model, recognizes detailed information regarding cloud infrastructure resources and correlation information from the input prompt, and transmits the recognized detailed information to the cloud infrastructure information generation unit.

[0018] In addition, the cloud infrastructure information generation unit can query and extract cloud infrastructure resources according to the text from a plurality of CSPs (Cloud Service Providers), map the cloud infrastructure resources according to the correlation information to establish correlations between the cloud infrastructure resources, and automatically convert the cloud infrastructure resources with established correlations into IaC (Infrastructure as Code) code to generate cloud infrastructure.

[0019] An artificial intelligence-based cloud infrastructure generation method according to another embodiment of the present invention comprises: a cloud infrastructure generation request step in which a cloud infrastructure generation request unit receives an image and uploads it to request the generation of cloud infrastructure; an image text conversion step in which an image text conversion unit converts the image into text using a pre-trained generative artificial intelligence model; a cloud infrastructure information generation step in which a cloud infrastructure information generation unit extracts cloud infrastructure resources based on the text and generates correlation information of the extracted cloud infrastructure resources; and a cloud infrastructure generation step in which a cloud infrastructure generation unit generates cloud infrastructure based on Infrastructure as Code (IaC) and cloud infrastructure according to the cloud infrastructure resources and correlation information.

[0020] Additionally, the cloud infrastructure architecture image providing unit may further include a cloud infrastructure architecture providing step of diagramming the architecture for the cloud infrastructure based on the code of the cloud infrastructure and providing the diagrammed cloud infrastructure architecture.

[0021] In addition, the cloud infrastructure creation request step described above can receive an image file in a raster graphic format provided by the user.

[0022] In addition, the image-to-text conversion step can recognize physical and virtual resource images representing cloud infrastructure resources, and block and line shapes representing the correlation between each resource, respectively, and convert the recognized information into text.

[0023] In addition, the cloud infrastructure creation request step described above can generate an image file in a vector graphic format using a drawing tool that provides physical and virtual resource image templates representing cloud infrastructure resources, and block and line shape templates for diagramming the correlations between each image.

[0024] Additionally, the cloud infrastructure detailed configuration request unit may further include a cloud infrastructure detailed configuration request step in which it receives a text-based prompt from a user based on a pre-trained language model, recognizes detailed information regarding cloud infrastructure resources and correlation information from the input prompt, and transmits the recognized detailed information to the cloud infrastructure information generation unit.

[0025] In addition, the cloud infrastructure information generation step can generate cloud infrastructure by querying and extracting cloud infrastructure resources according to the text from multiple CSPs (Cloud Service Providers), mapping the cloud infrastructure resources according to the correlation information to establish correlations between the cloud infrastructure resources, and automatically converting the correlated cloud infrastructure resources into Infrastructure as Code (IaC) code. Effects of the invention

[0026] According to the present invention, an AI-based cloud infrastructure creation device and method can be provided that allows for easy configuration of various cloud infrastructures and offers various options to users who wish to adopt and utilize the cloud, and supports safe cloud adoption by creating infrastructure based on Infrastructure as Code (IaC), which is advantageous for cloud infrastructure management, thereby preventing human errors and reducing management points. Brief explanation of the drawing

[0027] FIG. 1 is a schematic diagram illustrating the hardware configuration and overall operation of an artificial intelligence-based cloud infrastructure creation device according to one embodiment of the present invention. FIG. 2 is a block diagram showing the configuration of an artificial intelligence-based cloud infrastructure creation device according to one embodiment of the present invention. FIG. 3 is a diagram exemplarily illustrating the execution screen and operation process of an infrastructure resource drawing tool according to one embodiment of the present invention. FIG. 4 is a diagram illustrating the IaC conversion process through resource lookup / extraction and correlation information generation / setting according to an embodiment of the present invention. FIG. 5 is a diagram showing an example of cloud infrastructure resource extraction and correlation setting according to an embodiment of the present invention. FIG. 6 is a diagram showing the detailed operation process of an artificial intelligence-based cloud infrastructure creation device according to one embodiment of the present invention. FIG. 7 is a diagram exemplarily illustrating an IaC code automatic conversion screen of an infrastructure resource according to one embodiment of the present invention. FIG. 8 is a flowchart showing the configuration of an artificial intelligence-based cloud infrastructure creation method according to another embodiment of the present invention. Specific details for implementing the invention

[0028] The terms used in this specification will be briefly explained, and the invention will be described in detail.

[0029] The terms used in this invention have been selected based on currently widely used general terms, taking into account their functions within the invention; however, these terms may vary depending on the intent of those skilled in the art, case law, the emergence of new technologies, etc. Additionally, in specific cases, terms have been arbitrarily selected by the applicant, and in such cases, their meanings will be described in detail in the relevant description of the invention. Therefore, the terms used in this invention should be defined not merely by their names, but based on their meanings and the overall content of the invention.

[0030] When a part of a specification is described as "comprising" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components. Furthermore, terms such as "...part" or "module" as used in the specification refer to a unit that processes at least one function or operation, and this may be implemented in hardware or software, or as a combination of hardware and software.

[0031] Embodiments of the present invention are described below with reference to the attached drawings so that those skilled in the art can easily implement them. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals.

[0032] FIG. 1 is a schematic diagram illustrating the hardware configuration and overall operation of an artificial intelligence-based cloud infrastructure creation device according to one embodiment of the present invention.

[0033] Referring to FIG. 1, an artificial intelligence-based cloud infrastructure creation device (1000) according to one embodiment of the present invention may include at least one of a user terminal (10) and a cloud infrastructure creation server (20).

[0034] The above user terminal (10) is a terminal possessed by a user utilizing a cloud infrastructure service, and can receive an image (an image of the cloud infrastructure architecture requested by the user) from the user and upload it to a cloud infrastructure creation server (20) to request the creation of cloud infrastructure, and accordingly receive a diagrammed cloud infrastructure architecture image to allow the user to check visual information about the cloud infrastructure to be provided.

[0035] The user terminal (10) according to the present embodiment can be implemented in a form that applies the hardware configuration of computing devices such as a smartphone, smartpad, tablet PC, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), and Wibro (Wireless Broadband Internet) terminal.

[0036] The cloud infrastructure generation server (20) can convert an image uploaded by a user terminal (10) into text using a pre-trained generative artificial intelligence model, analyze the content of the converted text, query and extract cloud infrastructure resources from a CSP (Cloud Service Provider) based on the analyzed text information, generate correlation information of the extracted cloud infrastructure resources, generate cloud infrastructure based on Infrastructure as Code (IaC) according to the cloud infrastructure resources and correlation information, diagram the architecture of the cloud infrastructure based on the code of the generated cloud infrastructure, and provide the diagrammed cloud infrastructure architecture to the user terminal (10).

[0037] The cloud infrastructure creation server (20) according to the present embodiment may include program modules that perform various functions by being implemented in software using various forms of languages ​​such as C, C++, Java, Visual Basic, Visual C, etc., and may be implemented using web server programs provided in various ways according to operating systems such as DOS, Windows, Linux, Unix, Macintosh, Android, and iOS on general server hardware.

[0038] According to the present embodiment, the user terminal (10) and the cloud infrastructure creation server (20) can be connected via a network communication network, wherein the network communication network refers to a connection structure capable of exchanging information between each node, and examples include, but are not limited to, a 3GPP (3rd Generation Partnership Project) network, an LTE (Long Term Evolution) network, a WIMAX (World Interoperability for Microwave Access) network, the Internet, a LAN (Local Area Network), a Wireless LAN (Wireless Local Area Network), a WAN (Wide Area Network), a PAN (Personal Area Network), a Bluetooth network, a satellite broadcasting network, an analog broadcasting network, a DMB (Digital Multimedia Broadcasting) network, etc.

[0039] The configuration of the cloud infrastructure creation device (1000) according to the present embodiment will be described in more detail below with reference to the attached drawings.

[0040] FIG. 2 is a block diagram showing the configuration of an artificial intelligence-based cloud infrastructure creation device according to an embodiment of the present invention; FIG. 3 is a diagram exemplarily showing the execution screen and operation process of an infrastructure resource drawing tool according to an embodiment of the present invention; FIG. 4 is a diagram showing the process of IaC conversion through resource inquiry / extraction and correlation information generation / setting according to an embodiment of the present invention; FIG. 5 is a diagram showing an example of cloud infrastructure resource extraction and correlation setting according to an embodiment of the present invention; FIG. 7 is a diagram exemplarily showing an IaC code automatic conversion screen of an infrastructure resource according to an embodiment of the present invention; and FIG. 8 is a diagram showing the detailed operation process of an artificial intelligence-based cloud infrastructure creation device according to an embodiment of the present invention.

[0041] Referring to FIG. 2, an artificial intelligence-based cloud infrastructure creation device (1000) according to one embodiment of the present invention may include at least one of a cloud infrastructure creation request unit (100), an image text conversion unit (200), a cloud infrastructure information creation unit (300), a cloud infrastructure detailed setting request unit (400), a cloud infrastructure creation unit (500), and a cloud infrastructure architecture provision unit (600).

[0042] The above cloud infrastructure creation request unit (100) can request the creation of cloud infrastructure by receiving and uploading an image.

[0043] According to the present embodiment, the image uploaded by the user for a request to create cloud infrastructure is an image depicting cloud infrastructure resources and their correlations. It may be provided in various ways, such as an image drawn directly by the user on paper (applied after scanning), a photograph file of the image drawn on paper, or an image file created using a drawing program. It is preferable that the image also includes objects representing cloud infrastructure resources, the names (or identification information) of the objects, and inclusion and connection relationships between each object.

[0044] More specifically, the cloud infrastructure creation request unit (100) can receive an image file in a raster graphic format provided by the user.

[0045] For example, a user may request the creation of cloud infrastructure by directly drawing physical resource object images and / or virtual resource object images representing cloud infrastructure resources on paper, writing descriptions (names, descriptions, metadata, etc.) for each object image in adjacent locations, and creating an architecture image of the cloud infrastructure desired to be configured by drawing inclusion relationships or connection relationships between each object image with lines or blocks, and then creating a file by scanning or taking a picture of the image and uploading it to the image text conversion unit (200) of the cloud creation server (20) through the cloud infrastructure creation request unit (100).

[0046] Additionally, the cloud infrastructure creation request unit (100) can create an image file in a vector graphic format using a drawing tool that provides physical resource and virtual resource image templates representing cloud infrastructure resources and block and line shape templates for diagramming the correlation between each image.

[0047] For example, as illustrated in FIG. 3, when a drawing tool selects physical resource objects and / or virtual resource objects through a canvas and drags and drops the selected object images onto the canvas, descriptions (name, description, meta-information) for each dropped object image are automatically displayed in adjacent locations. Then, when a desired cloud infrastructure architecture is configured by selecting lines or / and blocks within the template, such as inclusion relationships or connection relationships between each object image, an image file containing the architecture is created and uploaded to the image text conversion unit (200) of the cloud creation server (20) through the cloud infrastructure creation request unit (100), thereby requesting the creation of cloud infrastructure.

[0048] The above image-to-text conversion unit (200) can convert an image (an architecture image uploaded by a user) into text using a pre-trained Generative Artificial Intelligence model.

[0049] More specifically, the image-text conversion unit (200) can identify and recognize physical resource and virtual resource images representing cloud infrastructure resources and block and line shapes representing correlations between each resource, respectively, and convert the identified and recognized information (cloud resources requiring infrastructure configuration, correlations between each resource) into text.

[0050] The generative AI model for this purpose is a model trained to convert images into text using an image encoder and an LLM decoder, and such a model is also referred to as image captioning. The LLM decoder (a language model based on GPT-2) plays the role of generating text by utilizing encoded vectors. The generative AI model according to the present embodiment may be structured to consist of image input, CNN (Inception V3), feature vector extraction (Embedding Vector), LSTM-based RNN (Text Decoder), and caption generation (Caption). Here, the image encoder plays the role of extracting image features through a CNN (encoding the image into meaningful feature vectors), and can encode images using a pre-trained CNN model such as 'Inception V3'. The LLM decoder plays the role of generating sentences using an LSTM (generating text by utilizing encoded vectors), and can input the extracted image feature vectors into the LSTM network. The LSTM generates words sequentially, and each word can be predicted based on the previous word and image features.

[0051] The above-mentioned cloud infrastructure information generation unit (300) can extract cloud infrastructure resources based on text (cloud resources requiring infrastructure configuration, correlations between each resource) converted through the image text conversion unit (200), and can generate and set correlation information of the extracted cloud infrastructure resources.

[0052] More specifically, the cloud infrastructure information generation unit (300) can query and extract cloud infrastructure resources based on text information from multiple CSPs (Cloud Service Providers), and map the cloud infrastructure resources based on correlation information between infrastructure resources to establish correlations between cloud infrastructure resources.

[0053] For example, as illustrated in FIG. 4, a user can query resources that they wish to configure in the architecture from among the text information, and these resources can be queried from a CSP, and the queried resources can be extracted for infrastructure configuration. The CSP according to the present embodiment may include, but is not limited to, Microsoft Azure, Amazon Web Services (AWS), Naver Cloud Service (Ncloud), etc. In this way, resources required for architecture configuration (e.g., Azure Management API, AWS SDK, API) can be extracted respectively.

[0054] Subsequently, necessary resources can be mapped according to the correlation between infrastructure resources included in the text information, and a mapping table can be generated as a result of the mapping. When mapping, resource setting values ​​can also be mapped by specification, image, network, and subnet interface, respectively, as shown in FIG. 5, and this resource mapping information can be generated in the form of a table.

[0055] The above cloud infrastructure detailed setting request unit (400) can receive a text-based prompt from a user based on a pre-trained language model, recognize detailed information regarding cloud infrastructure resources and correlation information from the input prompt, and transmit the recognized detailed information to the cloud infrastructure information generation unit (300).

[0056] For example, as shown in FIG. 6, when text such as 'Create an Azure 3-Tier architecture!' is entered through the language model, a prompt for identifying detailed resources and setting detailed relationships between resources is recognized from the entered text, and after confirming the recognized prompt to the user, the prompt (in text form) is entered into the cloud infrastructure information generation unit (300) so that detailed settings for resources and relationships are performed.

[0057] The above-mentioned cloud infrastructure detailed setting request unit (400) can be configured not only to directly input text into the cloud infrastructure information generation unit (300), but also to directly input text required for detailed setting into the image text conversion unit (200).

[0058] Accordingly, the cloud infrastructure information generation unit (300) can extract and set the cloud infrastructure components that the user wants to configure by combining the text converted through the image text conversion unit (200) and the text recognized through the cloud infrastructure detailed setting request unit (400).

[0059] The above cloud infrastructure creation unit (500) can create cloud infrastructure based on IaC (Infrastructure as Code) according to cloud infrastructure resources and correlation information.

[0060] For example, as illustrated in Fig. 7, a new cloud infrastructure can be configured by automatically converting each correlated cloud infrastructure resource into Infrastructure as Code (IaC) code. Infrastructure as Code (IaC) allows for the management and provisioning of infrastructure through code rather than through manual processes. When using IaC during cloud infrastructure construction, configuration files containing infrastructure specifications are generated, making it easy to edit and deploy the infrastructure configuration. Additionally, IaC ensures that the same environment is provisioned every time and supports configuration management by codifying and documenting configuration specifications, thereby preventing temporary changes without documentation.

[0061] The cloud infrastructure architecture providing unit (600) can diagram the architecture of the cloud infrastructure created by the cloud infrastructure creation unit (500) based on the code of the cloud infrastructure, and provide the diagrammed cloud infrastructure architecture to the user terminal (10). Accordingly, the user terminal (10) can compare the architecture image uploaded by the user with the architecture image provided by the cloud infrastructure creation server (20) to verify whether the requested infrastructure is properly configured.

[0062] FIG. 8 is a flowchart showing the configuration of an artificial intelligence-based cloud infrastructure creation method according to another embodiment of the present invention.

[0063] Referring to FIG. 8, an artificial intelligence-based cloud infrastructure creation method (S1000) according to one embodiment of the present invention may include at least one of a cloud infrastructure creation request step (S100), an image text conversion step (S200), a cloud infrastructure information creation step (S300), a cloud infrastructure detailed setting request step (S400), a cloud infrastructure creation step (S500), and a cloud infrastructure architecture provision step (S600).

[0064] In the above cloud infrastructure creation request step (S100), an image can be received as input and uploaded to request the creation of cloud infrastructure.

[0065] According to the present embodiment, the image uploaded by the user for a request to create cloud infrastructure is an image depicting cloud infrastructure resources and their correlations. It may be provided in various ways, such as an image drawn directly by the user on paper (applied after scanning), a photograph file of the image drawn on paper, or an image file created using a drawing program. It is preferable that the image also includes objects representing cloud infrastructure resources, the names (or identification information) of the objects, and inclusion and connection relationships between each object.

[0066] More specifically, in the cloud infrastructure creation request step (S100), an image file in a raster graphic format provided by the user can be received.

[0067] For example, a user may request the creation of cloud infrastructure by directly drawing physical resource object images and / or virtual resource object images representing cloud infrastructure resources on paper, writing descriptions (names, descriptions, metadata, etc.) for each object image in adjacent locations, and creating an architecture image of the cloud infrastructure desired to be configured by drawing inclusion relationships or connection relationships between each object image with lines or blocks, and then creating a file by scanning or taking a picture of the image and uploading it to the image text conversion step (S200) of the cloud creation server (20) through the cloud infrastructure creation request step (S100).

[0068] Additionally, in the cloud infrastructure creation request step (S100), an image file in a vector graphic format can be created using a drawing tool that provides physical and virtual resource image templates representing cloud infrastructure resources, and block and line shape templates for diagramming the correlation between each image.

[0069] For example, as illustrated in FIG. 3, when a drawing tool selects physical resource objects and / or virtual resource objects through a canvas and drags and drops the selected object images onto the canvas, descriptions (name, description, meta-information) for each dropped object image are automatically displayed in adjacent locations. Then, when a desired cloud infrastructure architecture is configured by selecting lines or / and blocks within the template, such as inclusion relationships or connection relationships between each object image, an image file containing the architecture is created and uploaded to the image text conversion step (S200) of the cloud creation server (20) through the cloud infrastructure creation request step (S100), thereby requesting the creation of cloud infrastructure.

[0070] In the above image-to-text conversion step (S200), an image (an architecture image uploaded by a user) can be converted into text using a pre-trained Generative Artificial Intelligence model.

[0071] More specifically, in the image-to-text conversion step (S200), images of physical and virtual resources representing cloud infrastructure resources and block and line shapes representing correlations between each resource are identified and recognized, respectively, and the identified and recognized information (cloud resources requiring infrastructure configuration, correlations between each resource) can be converted into text.

[0072] The generative AI model for this purpose is a model trained to convert images into text using an image encoder and an LLM decoder, and such a model is also referred to as image captioning. The LLM decoder (a language model based on GPT-2) plays the role of generating text by utilizing encoded vectors. The generative AI model according to the present embodiment may be structured to consist of image input, CNN (Inception V3), feature vector extraction (Embedding Vector), LSTM-based RNN (Text Decoder), and caption generation (Caption). Here, the image encoder plays the role of extracting image features through a CNN (encoding the image into meaningful feature vectors), and can encode images using a pre-trained CNN model such as 'Inception V3'. The LLM decoder plays the role of generating sentences using an LSTM (generating text by utilizing encoded vectors), and can input the extracted image feature vectors into the LSTM network. The LSTM generates words sequentially, and each word can be predicted based on the previous word and image features.

[0073] In the above cloud infrastructure information generation step (S300), cloud infrastructure resources are extracted based on the text (cloud resources requiring infrastructure configuration, correlations between each resource) converted through the image text conversion step (S200), and correlation information of the extracted cloud infrastructure resources can be generated and set.

[0074] More specifically, in the cloud infrastructure information generation step (S300), cloud infrastructure resources based on text information are queried and extracted from multiple CSPs (Cloud Service Providers), and correlations between cloud infrastructure resources can be established by mapping the cloud infrastructure resources according to correlation information between infrastructure resources.

[0075] For example, as illustrated in FIG. 4, a user can query resources that they wish to configure in the architecture from among the text information, and these resources can be queried from a CSP, and the queried resources can be extracted for infrastructure configuration. The CSP according to the present embodiment may include, but is not limited to, Microsoft Azure, Amazon Web Services (AWS), Naver Cloud Service (Ncloud), etc. In this way, resources required for architecture configuration (e.g., Azure Management API, AWS SDK, API) can be extracted respectively.

[0076] Subsequently, necessary resources can be mapped according to the correlation between infrastructure resources included in the text information, and a mapping table can be generated as a result of the mapping. When mapping, resource setting values ​​can also be mapped by specification, image, network, and subnet interface, respectively, as shown in FIG. 5, and this resource mapping information can be generated in the form of a table.

[0077] In the above cloud infrastructure detailed configuration request step (S400), a text-based prompt is received from a user based on a pre-trained language model, detailed information regarding cloud infrastructure resources and correlation information is recognized from the input prompt, and the recognized detailed information can be transmitted to the cloud infrastructure information generation step (S300).

[0078] For example, as illustrated in Fig. 6, when text such as 'Create Azure 3-Tier Architecture!' is entered through the language model, a prompt for identifying detailed resources and setting detailed relationships between resources is recognized from the entered text, and after confirming the recognized prompt to the user, the prompt (in text form) is entered into the cloud infrastructure information creation step (S300) so that detailed settings for resources and relationships are performed.

[0079] In the above cloud infrastructure detailed setting request step (S400), the method of directly inputting text into the cloud infrastructure information generation step (S300) can be configured to directly input the text required for detailed settings into the image text conversion step (S200).

[0080] Accordingly, in the cloud infrastructure information generation step (S300), the text converted through the image text conversion step (S200) and the text recognized through the cloud infrastructure detailed setting request step (S400) can be combined to extract and set the cloud infrastructure components that the user wishes to configure.

[0081] In the above cloud infrastructure creation step (S500), cloud infrastructure can be created based on Infrastructure as Code (IaC) according to cloud infrastructure resources and correlation information.

[0082] For example, as illustrated in Fig. 7, new cloud infrastructure can be created by automatically converting each correlated cloud infrastructure resource into Infrastructure as Code (IaC) code. Infrastructure as Code (IaC) allows infrastructure to be managed and provisioned through code rather than through manual processes. When using IaC to build cloud infrastructure, configuration files containing infrastructure specifications are generated, making it easy to edit and deploy the infrastructure configuration. Additionally, IaC ensures that the same environment is provisioned every time and supports configuration management by codifying and documenting configuration specifications, thereby preventing temporary changes without documentation.

[0083] In the cloud infrastructure architecture provision step (S600), the architecture for the cloud infrastructure created in the cloud infrastructure creation step (S500) is diagrammed based on the code of the cloud infrastructure, and the diagrammed cloud infrastructure architecture can be provided to the user terminal (10). Accordingly, the user terminal (10) can compare the architecture image uploaded by the user with the architecture image provided by the cloud infrastructure creation server (20) to verify whether the requested infrastructure is properly configured.

[0084] The AI-based cloud infrastructure generating device described above may be implemented as a hardware component, a software component, and / or a combination of a hardware component and a software component. For example, the device and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions.

[0085] An AI-based cloud infrastructure generating device may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the AI-based cloud infrastructure generating device may access, store, manipulate, process, and generate data in response to the execution of software. For ease of understanding, the AI-based cloud infrastructure generating device may be described as being used as a single unit, but a person of ordinary knowledge in the relevant technical field will know that the AI-based cloud infrastructure generating device may include multiple processing elements and / or multiple types of processing elements.

[0086] For example, an AI-based cloud infrastructure creation device may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible. Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure the processing device to operate as desired or command the processing device independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave to be interpreted by the processing device or to provide instructions or data to the processing device. Software may be distributed over networked computer systems, stored or executed in a distributed manner, or stored on one or more computer-readable recording media.

[0087] The above description is merely one embodiment for implementing the artificial intelligence-based cloud infrastructure creation device and method according to the present invention. The present invention is not limited to the above embodiment, and the technical spirit of the present invention extends to the scope in which various modifications can be made by anyone with ordinary knowledge in the field to which the invention belongs, without departing from the gist of the invention as claimed in the following patent claims. Explanation of the symbols

[0088] 1000: AI-based cloud infrastructure creation device 100: Cloud infrastructure creation request 200: Image-to-Text Conversion Section 300: Cloud Infrastructure Information Generation Section 400: Cloud Infrastructure Detailed Configuration Request 500: Cloud Infrastructure Creation Section 600: Cloud Infrastructure Architecture Provider S1000: Method for Creating AI-Based Cloud Infrastructure S100: Cloud infrastructure creation request step S200: Image-to-text conversion step S300: Cloud infrastructure information generation step S400: Cloud infrastructure detailed configuration request step S500: Cloud infrastructure creation phase 600: Cloud Infrastructure Architecture Provider Phase

Claims

Claim 1 A cloud infrastructure creation request unit that receives and uploads an image to request the creation of cloud infrastructure; an image-to-text conversion unit that converts the image into text using a pre-trained generative artificial intelligence model; and a cloud infrastructure information generation unit that extracts cloud infrastructure resources based on the text and generates correlation information of the extracted cloud infrastructure resources.The cloud infrastructure generation unit includes a cloud infrastructure generation unit that generates cloud infrastructure based on cloud infrastructure resources and correlation information based on Infrastructure as Code (IaC). The cloud infrastructure generation request unit generates an image file in a vector graphic format using a drawing tool that provides physical and virtual resource image templates representing cloud infrastructure resources and block and line shape templates for diagramming the correlations between each image. After receiving a selection from a user of at least one of physical resource objects and virtual resource objects through a canvas, when the selected object image is dragged and dropped onto the canvas, descriptive information (name, description, meta-information) for each dropped object image is automatically displayed at a location adjacent to the object image. The cloud infrastructure architecture image that the user wishes to configure is generated by selecting at least one of line and block shapes within the template to represent the inclusion and connection relationships between each object image. The image text conversion unit uses a generative artificial intelligence model trained on a model that converts the cloud infrastructure architecture image generated through the cloud infrastructure generation request unit into text, to identify and recognize the physical and virtual resource images representing cloud infrastructure resources and the block and line shapes representing the correlations between each resource, respectively, and the identified and recognized information (infrastructure configuration An AI-based cloud infrastructure generation device characterized by converting necessary cloud infrastructure resources and correlations between each resource into text, and the generative AI model using an image encoder that extracts image features by encoding an image into a feature vector through a Convolutional Neural Network (CNN), and a Large Language Model (LM) decoder that generates sentence-form text from the encoded vector through a Long Short-Term Memory (LSTM). Claim 2 An AI-based cloud infrastructure generation device according to claim 1, further comprising a cloud infrastructure architecture providing unit that diagrams an architecture for the cloud infrastructure based on the code of the cloud infrastructure and provides the diagrammed cloud infrastructure architecture. Claim 3 In claim 1, the cloud infrastructure creation request unit is characterized by receiving an image file in a raster graphic format provided by a user, in an AI-based cloud infrastructure creation device. Claim 4 delete Claim 5 delete Claim 6 An artificial intelligence-based cloud infrastructure generation device according to claim 1, further comprising a cloud infrastructure detailed setting request unit that receives a text-based prompt from a user based on a pre-trained language model, recognizes detailed information regarding cloud infrastructure resources and correlation information from the input prompt, and transmits the recognized detailed information to the cloud infrastructure information generation unit. Claim 7 An AI-based cloud infrastructure generation device according to claim 1, wherein the cloud infrastructure information generation unit queries and extracts cloud infrastructure resources according to the text from a plurality of CSPs (Cloud Service Providers), maps the cloud infrastructure resources according to the correlation information to establish correlations between the cloud infrastructure resources, and automatically converts the cloud infrastructure resources with established correlations into IaC (Infrastructure as Code) code to generate cloud infrastructure. Claim 8 A cloud infrastructure creation request step in which a cloud infrastructure creation request unit receives an image as input and uploads it to request the creation of cloud infrastructure; an image text conversion step in which an image text conversion unit converts the image into text using a pre-trained generative artificial intelligence model; and a cloud infrastructure information generation step in which a cloud infrastructure information generation unit extracts cloud infrastructure resources based on the text and generates correlation information of the extracted cloud infrastructure resources.The cloud infrastructure creation unit includes a cloud infrastructure creation step that creates cloud infrastructure based on Infrastructure as Code (IaC) according to cloud infrastructure resources and correlation information, and the cloud infrastructure creation request step creates an image file in a vector graphic format using a drawing tool that provides physical resource and virtual resource image templates representing cloud infrastructure resources and block and line shape templates for diagramming the correlation between each image, wherein at least one of the physical resource object and virtual resource object is selected by the user through a canvas, and when the selected object image is dragged and dropped onto the canvas, descriptive information (name, description, meta-information) for each dropped object image is automatically displayed at a location adjacent to the object image, and the inclusion and connection relationships between each object image are created by the user through the selection of at least one of the line and block shapes within the template, thereby creating a cloud infrastructure architecture image that the user wishes to configure, and the image text conversion step uses a generative artificial intelligence model trained on a model that converts the cloud infrastructure architecture image created through the cloud infrastructure creation request step into text, to identify and recognize the physical resource and virtual resource images representing cloud infrastructure resources and the block and line shapes representing the correlation between each resource, respectively, and the identified and recognized An AI-based method for generating cloud infrastructure, characterized by converting information (cloud infrastructure resources requiring infrastructure configuration and correlations between each resource) into text, and the generative AI model using an image encoder that extracts image features by encoding an image into feature vectors through a Convolutional Neural Network (CNN), and a Large Language Model (LM) decoder that generates sentence-form text from the encoded vectors through a Long Short-Term Memory (LSTM). Claim 9 A method for creating an AI-based cloud infrastructure according to claim 8, characterized in that the cloud infrastructure architecture image providing unit further includes a cloud infrastructure architecture providing step of diagramming an architecture for the cloud infrastructure based on the code of the cloud infrastructure and providing the diagrammed cloud infrastructure architecture. Claim 10 In claim 8, the cloud infrastructure creation request step is characterized by receiving an image file in a raster graphic format provided by a user, in an artificial intelligence-based cloud infrastructure creation method. Claim 11 delete Claim 12 delete Claim 13 An AI-based cloud infrastructure generation method according to claim 8, further comprising a cloud infrastructure detailed configuration request unit receiving a text-based prompt from a user based on a pre-trained language model, recognizing detailed information regarding cloud infrastructure resources and correlation information from the input prompt, and transmitting the recognized detailed information to the cloud infrastructure information generation unit. Claim 14 In claim 8, the cloud infrastructure information generation step comprises querying and extracting cloud infrastructure resources according to the text from a plurality of CSPs (Cloud Service Providers), mapping the cloud infrastructure resources according to the correlation information to establish correlations between the cloud infrastructure resources, and automatically converting the cloud infrastructure resources with established correlations into Infrastructure as Code (IaC) code to generate cloud infrastructure, characterized by an AI-based cloud infrastructure generation method. Claim 15 A computer-readable recording medium having a program for executing a method for creating an artificial intelligence-based cloud infrastructure according to any one of claims 8 to 10 and 13 to 14.

Images