Electronic apparatus for determining collaboration possibility between enterprises based on rag and collaboration enterprise matching method using the same

Abstract

The electronic device of the present disclosure analyzes input data including a user query, retrieves relevant data corresponding to said input data from a corporate information vector index in which at least one of anonymized corporate technical information, certification status, and equipment list is vectorized and stored, and generates answer data based on said relevant data to provide an answer to said user query, and can perform at least one subsequent task based on the context of said query and said answer using an AI agent. said at least one subsequent task may include determining the possibility of collaboration between a first company within a first industrial cluster and at least one other company within said first industrial cluster based on at least one of said input data, said relevant data, and said answer data.

Description

Technology Field

[0001] The present disclosure relates to a technology for matching companies capable of collaboration, and more specifically, to an electronic device that generates an answer to a user's query based on RAG and determines the possibility of collaboration between companies using an AI agent, and a method for matching companies capable of collaboration using the same. Background Technology

[0002] Recently, data-driven analytical technologies for exploring inter-company collaboration opportunities and assessing collaboration potential are becoming increasingly widespread. In particular, attempts are being made to analyze corporate technical information, financial status, and certification history using artificial intelligence, data mining, and recommendation algorithms to estimate collaboration possibilities. However, existing methods are primarily centered on individual databases or corporate information within specific industries, making it difficult to extend them to an integrated analysis of collaboration potential across entire industries.

[0003] Conventional technology relies on methods that simply compare corporate attribute data or calculate similarity, making it difficult to effectively reflect multidimensional and unstructured corporate data. Furthermore, because the results of assessing collaboration potential in conventional technology are fragmentary and highly uncertain, it is difficult to provide highly reliable evaluations that lead to actual collaboration. Additionally, conventional technology has limited data sources and lacks sufficient features to guarantee de-identification or anonymity, which can lead to data security and reliability issues when utilizing sensitive corporate information.

[0004] Therefore, there is a growing need for new technological means to more accurately and reliably assess collaboration potential among various companies within an industrial cluster. In particular, there is an increasing demand for methods that can efficiently search and analyze vast amounts of corporate data while ensuring data security and anonymity, and rapidly identify collaboration opportunities. The problem to be solved

[0005] One objective of the present disclosure is to provide an electronic device capable of determining the potential for collaboration between companies with high accuracy by efficiently integrating and analyzing various company data within an industrial cluster, and a method for matching collaborating companies using the same.

[0006] Another objective of the present disclosure is to provide an electronic device and a method for matching collaborative companies using the same, which enables a sophisticated evaluation of collaboration potential beyond simple attribute comparison by reflecting multidimensional data attributes through RAG-based analysis and the utilization of AI agents.

[0007] Another objective of the present disclosure is to provide an electronic device and a method for matching collaborating companies using the same, which enables the safe utilization of sensitive data by processing corporate information in a de-identified state, and simultaneously secures trust in data-providing companies and data security during the collaboration search process through an anonymous-based intermediary protocol.

[0008] Another objective of the present disclosure is to provide an electronic device and a method for matching collaborative companies using the same, which increases the likelihood of leading to actual collaboration by providing the results of the collaboration possibility assessment as a highly reliable indicator and enables the effective discovery of new partnership opportunities within an industrial cluster.

[0009] However, the problems that this disclosure aims to solve are not limited to those mentioned above, and other unmentioned problems will be clearly understood by a person skilled in the art from the description below. means of solving the problem

[0010] An electronic device according to the present disclosure for achieving the aforementioned technical objectives, which generates an answer to a user's query based on RAG and determines the possibility of inter-company collaboration using an AI agent, may include a memory that stores at least one instruction and at least one processor that executes said at least one instruction. The at least one processor may analyze input data including the user's query, search for relevant data corresponding to the input data from a corporate information vector index in which at least one of anonymized corporate technical information, certification status, and equipment list is vectorized and stored, provide an answer to the user's query by generating answer data based on said relevant data, and perform at least one subsequent task based on the context of said query and said answer using an AI agent. The at least one subsequent task may include determining the possibility of collaboration between a first company within a first industrial cluster and at least one other company within the first industrial cluster based on at least one of said input data, said relevant data, and said answer data.

[0011] In one embodiment, the at least one processor may provide a mediation protocol that anonymously mediates collaboration between at least one other company matched with the first company.

[0012] In one embodiment, the at least one processor can tokenize input data using a morphological analysis module, extract entities and attributes from a query included in the input data, and determine the intent and subject of the query based on session metadata.

[0013] In one embodiment, the at least one processor can convert a query included in the input data into a high-dimensional query vector using an embedding model and set a filter parameter including at least one of an industry sector, technology category, certification level, and facility size associated with the query vector.

[0014] In one embodiment, the at least one processor can perform a nearest neighbor search for a non-identifying corporate data vector mapped to the same vector space as the query vector and extract at least one corporate anonymous identifier based on a similarity score.

[0015] In one embodiment, the at least one processor can search the private collection and the public collection in parallel and sort the private collection and the public collection in a preset sorting method to generate a corporate information list.

[0016] In one embodiment, the relevant data may include at least one of an anonymized enterprise technical description, equipment specification information, certification history information, production capacity indicators, quality control indicators, process standard documents, safety compliance records, maintenance history, supply chain linkage information, geographic location information, timestamps, and source metadata.

[0017] In one embodiment, the at least one processor calculates at least one of industry suitability, technical similarity, authentication reliability, recency, and source reliability for the relevant data, and can readjust the result ranking of the relevant data based on the calculation result.

[0018] In one embodiment, the at least one processor can construct an extension context by combining relevant data and a query, and generate answer data including at least one of a collaboration candidate list, a comparison analysis table, and a suitability evaluation report by calling an LLM API.

[0019] In one embodiment, the at least one processor converts the answer data into at least one of a natural language format, a structured data (JSON) format, and a visualization data format, and can transmit the converted answer data in a format optimized for the user environment.

[0020] In one embodiment, the at least one processor may sequentially perform collaboration matching candidate derivation, comparative analysis, and anonymous connection request when the collaboration possibility is greater than or equal to the reference matching score, and may perform alternative industry cluster search or new partner recommendation when the collaboration possibility is less than the reference matching score.

[0021] In one embodiment, the at least one processor transmits an initial collaboration request between companies matched based on an anonymous ID, exchanges real-name data only when approval of the collaboration request is made between companies, and can provide the execution status and result log of at least one subsequent task.

[0022] In one embodiment, the at least one processor creates a secure collaboration channel between a first company and another company, transmits a schedule draft, a task list, and a data exchange policy through the secure collaboration channel, and can store the document exchange history and approval log of the secure collaboration channel in conjunction with matching data including at least one of a search basis, a technical tag, and an inter-company complementarity analysis generated during the collaboration matching process.

[0023] A collaborative enterprise matching method according to the present disclosure for achieving the aforementioned technical objectives may include: a step of analyzing input data including a user query; a step of retrievaling relevant data corresponding to said input data from an enterprise information vector index in which at least one of anonymized enterprise technical information, certification status, and facility list is vectorized and stored; a step of providing an answer to said user query by generating answer data based on said relevant data; and a step of performing at least one subsequent task based on the context of said query and said answer using an AI agent. The at least one subsequent task may include determining the possibility of collaboration between a first enterprise within a first industry cluster and at least one other enterprise within said first industry cluster based on at least one of said input data, said relevant data, and said answer data.

[0024] In one embodiment, the step of performing the at least one subsequent task may include providing an anonymous-based mediation protocol that mediates collaboration between the first company and the at least one other company matched with the first company.

[0025] In one embodiment, the step of performing the at least one subsequent task may include the step of sequentially performing collaboration matching candidate derivation, comparative analysis, and anonymous connection request when the collaboration possibility is greater than or equal to the reference matching score, and the step of performing alternative industry cluster search or new partner recommendation when the collaboration possibility is less than the reference matching score.

[0026] In one embodiment, the step of performing the at least one subsequent task may include the step of transmitting an initial collaboration request between companies matched based on an anonymous ID, the step of exchanging real-name data only when approval of the collaboration request is made between companies, and the step of providing the execution status and result log of the at least one subsequent task.

[0027] In one embodiment, the step of performing at least one subsequent task may include creating a secure collaboration channel between the first company and the first company, transmitting a schedule draft, a task list, and a data exchange policy through the secure collaboration channel, and storing the document exchange history and approval log of the secure collaboration channel in conjunction with matching data including at least one of a search basis, a technical tag, and an inter-company complementarity analysis generated during the collaboration matching process.

[0028] In addition to this, a computer program stored on a computer-readable recording medium for implementing the present disclosure may be further provided.

[0029] In addition to this, a computer-readable recording medium for recording a computer program for implementing the present disclosure may be further provided. Effects of the invention

[0030] According to the aforementioned means for solving the problem of the present disclosure, the electronic device and the collaborative enterprise matching method using the same of the present disclosure can determine the possibility of collaboration between companies with high accuracy by efficiently integrating and analyzing various company data within an industrial cluster. Since the electronic device and the collaborative enterprise matching method using the same of the present disclosure can reflect multidimensional data attributes through RAG-based analysis and the utilization of AI agents, it becomes possible to evaluate collaboration possibilities in a sophisticated manner that was lacking in conventional technology. Therefore, the electronic device and the collaborative enterprise matching method using the same of the present disclosure can significantly improve the efficiency of the process of discovering collaboration between companies.

[0031] Furthermore, the electronic device of the present disclosure and the collaborative company matching method using the same allow for the safe utilization of sensitive data because company information is processed in a de-identified state. The electronic device of the present disclosure and the collaborative company matching method using the same can simultaneously secure data security and privacy during the collaboration search process while maintaining the trust of data-providing companies through an anonymous-based intermediary protocol. Therefore, companies within an industrial cluster can review the possibility of collaboration with other companies without fear of data leakage.

[0032] Furthermore, the electronic device of the present disclosure and the collaborative company matching method using the same can provide the results of the assessment of collaboration potential as highly reliable indicators, thereby increasing the likelihood of leading to actual collaboration. Since the electronic device of the present disclosure and the collaborative company matching method using the same can effectively match candidate companies for collaboration based on the results of the collaboration potential assessment, new partnership opportunities within an industrial cluster can be secured more quickly and stably.

[0033] The effects of the present disclosure are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description below. Brief explanation of the drawing

[0034] FIG. 1 is a drawing showing a system (100) in which an industrial cluster including an electronic device of the present invention and a plurality of companies is connected. FIG. 2 is a diagram showing the block configuration of the electronic device of the present invention. FIG. 3 is a diagram showing the system architecture and data processing flow of the electronic device of the present invention. FIG. 4 is a flowchart illustrating the operation of the electronic device of the present invention. FIG. 5 is a conceptual diagram illustrating the operation of matching companies based on the collaboration possibilities of the present invention. FIG. 6 is a flowchart illustrating the operation of matching companies based on the collaboration possibilities of the present invention. FIG. 7 is a diagram showing a mediation protocol in which the electronic device of the present invention mediates between a plurality of matched companies. Specific details for implementing the invention

[0035] Hereinafter, embodiments are described in detail with reference to the attached drawings. However, various modifications may be made to the embodiments, and thus the scope of the patent application is not limited or restricted by these embodiments. It should be understood that all modifications, equivalents, and substitutions to the embodiments are included within the scope of the rights.

[0036] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, the embodiments are not limited to the specific disclosed forms, and the scope of this specification includes modifications, equivalents, or substitutions that fall within the technical concept.

[0037] Terms such as "first" or "second" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0038] Furthermore, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly and specifically defined otherwise. In certain cases, terms have been selected at the applicant's discretion, and in such cases, their meanings will be described in detail in the relevant explanatory sections. Accordingly, terms used in this disclosure should be defined not merely by their names, but based on their meanings and the content throughout this disclosure.

[0039] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or joined to that other component, or that there may be other components in between.

[0040] The terms used in the embodiments are for illustrative purposes only and should not be interpreted as intended to be limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0041] Throughout this specification, when a part 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, the singular form used in this specification includes the plural form unless specifically stated otherwise in the text. Additionally, the expression "at least one of a, b, and c" described throughout this specification may encompass 'a alone,' 'b alone,' 'c alone,' 'a and b,' 'a and c,' 'b and c,' or 'a, b, and c all.'

[0042] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the embodiments pertain. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0043] Additionally, terms such as “part,” “module,” etc., as described in this specification refer to a unit that processes at least one function or operation, which may be implemented in hardware or software, or a combination of hardware and software. Furthermore, embodiments of the present disclosure may be represented by functional block configurations and various processing steps. These functional blocks may be implemented by various numbers of hardware and / or software configurations that execute specific functions. For example, embodiments of the present disclosure may employ integrated circuit configurations such as memory, processing, logic, look-up tables, etc., which can execute various functions under the control of one or more microprocessors or other control devices.

[0044] Each block of the process flow diagrams attached to this specification and combinations of the flow diagrams may be executed by computer program instructions. Since these computer program instructions may be loaded into the processor of a general-purpose computer, a computer for special purposes, or other programmable data processing equipment, the instructions executed through the processor of the computer or other programmable data processing equipment create means for performing the functions described in the flow diagram block(s).

[0045] These computer program instructions may be stored in computer-available or computer-readable memory that can be directed toward a computer or other programmable data processing equipment to implement a function in a specific way, and the instructions stored in said computer-available or computer-readable memory may also produce a manufactured item containing instruction means that performs the function described in the flowchart block(s).

[0046] Since computer program instructions can be loaded onto a computer or other programmable data processing equipment, instructions that perform 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 the flowchart block(s).

[0047] Additionally, each block may represent a module, segment, or part of code containing one or more executable instructions for executing a specified logical function(s). Furthermore, in some alternative execution examples, the functions mentioned in the blocks may occur out of order. For instance, two blocks described in succession may actually be executed substantially simultaneously, or the blocks may be executed in reverse order according to their corresponding functions.

[0048] In addition, when describing with reference to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the embodiments, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the embodiments, such detailed description is omitted.

[0049] FIG. 1 is a drawing showing a system (100) in which an industrial cluster including an electronic device of the present invention and a plurality of companies is connected.

[0050] Referring to FIG. 1, an industrial cluster may be a collection of companies in which multiple companies belonging to various industrial groups are mutually distributed. An industrial cluster may refer to a physical or virtual network in which resources and technologies are concentrated among companies, and individual companies within the industrial cluster may be companies that possess information and capabilities specialized in a specific industrial field.

[0051] For example, an industrial cluster may include companies from different fields, such as the robotics industry, medical device manufacturing, construction, information technology (IT) services, pharmaceutical and biotechnology industries, automotive industry, aerospace industry, manufacturing, distribution, logistics and transportation, financial services, educational services, energy and power industries, agriculture and food processing industries, and entertainment and content production industries. At least one company within the industrial cluster may possess corporate information directly related to industrial activities, such as technical information, facility lists, and certification status. For example, companies within the industrial cluster may be independent entities producing specific products or providing services, and may form cooperative or competitive relationships with one another.

[0052] The electronic device can communicate with the industrial cluster and may be a core processing unit comprising a Retrieval-Augmented Generation (RAG)-based computation module or an AI agent control module. Connections between the electronic device and enterprises within the industrial cluster can be formed based on data flow, and query analysis results, retrieved relevant data, generated collaboration proposals, and subsequent mediation requests can be exchanged through these connections. For example, the electronic device can connect with each enterprise via a network to perform functions such as query processing and searching for enterprise data, determining collaboration possibilities, and mediating between enterprises.

[0053] The electronic device can collect anonymized data from multiple companies and manage it as a vector index, enabling efficient retrieval of company information in response to queries. For example, the electronic device can collect data from individual companies within an industrial cluster from which direct identifying information has been removed, and perform the role of exploring collaboration possibilities based on that data. Furthermore, the electronic device can evaluate collaboration potential between multiple companies, recommend collaboration partners, and even carry out collaboration mediation procedures if necessary.

[0054] FIG. 2 is a diagram showing the block configuration of the electronic device (200) of the present invention.

[0055] Referring to FIG. 2, the electronic device communicating with the industrial cluster in FIG. 1 may be implemented as an electronic device (200). For example, the electronic device (200) may include at least one of a server PC in an in-house or edge environment, a public or private cloud server, and a database.

[0056] In one embodiment, the electronic device (200) may include an input / output interface (210), a memory (220), a processor (230), and a communication interface (240). However, the present disclosure is not limited thereto. The electronic device (200) may be configured with some of the components shown in FIG. 2 omitted, or may be configured to include other components in addition to the components shown in FIG. 2.

[0057] In one embodiment, the input / output interface (210), memory (220), processor (230), and communication interface (240) may each be physically / electrically connected to each other.

[0058] In one embodiment, the electronic device (200) may be connected to various types of external devices through an input / output interface (210). In one embodiment, the input / output interface (210) may include at least one of a wired / wireless headset port, an external charger port, a wired / wireless data port, a memory card port, a port for connecting a device equipped with an identification module (SIM), an audio I / O (Input / Output) port, or a video I / O (Input / Output) port. In one embodiment, the input / output interface (210) may include a USB (Universal Serial Bus), HDMI (High Definition Multimedia Interface), or DVI (Digital Visual Interface), etc.

[0059] In one embodiment, the memory (220) may store data used in the electronic device (200). In one embodiment, the memory (220) may store instructions, programs, or modules for the operation of the processor (230).

[0060] In one embodiment, the memory (220) may include at least one type of storage medium among a flash memory type, a hard disk type, an SSD type (Solid State Disk type), an SSD type (Silicon Disk Drive type), a multimedia card micro type, a card type memory (e.g., SD or XD memory, etc.), RAM (random access memory; RAM), SRAM (static random access memory), ROM (read-only memory; ROM), EEPROM (electrically erasable programmable read-only memory), PROM (programmable read-only memory), magnetic memory, and an optical disk.

[0061] In one embodiment, the memory (220) may include user information and components displayed on the user interface.

[0062] In one embodiment, the processor (230) may include a general-purpose processor such as a CPU (Central Processing Unit), AP (Application Processor), DSP (Digital Signal Processor), or a neural network processing processor such as an NPU (Neural Processing Unit). In one embodiment, the processor (230) may be divided by one or more processors to perform operations. In one embodiment, the processor (230) may control the operation of the electronic device (200). The processor (230) may control the operation of the electronic device (200) according to instructions stored in memory (220).

[0063] In one embodiment, the processor (230) may be implemented as a memory that stores data for an algorithm or a program that reproduces the algorithm for controlling the operation of components within the electronic device (200) of the present disclosure, and as a processor that performs the aforementioned operation using the data stored in the memory. In this case, the memory and the processor may each be implemented as separate chips. Alternatively, the memory and the processor may be implemented as a single chip.

[0064] In addition, the processor may control one or a combination of the components described above in order to implement various embodiments according to the present disclosure through the electronic device (200).

[0065] In one embodiment, the communication interface (240) may include one or more components that enable communication between an external server or an external electronic device and the electronic device (200). In one embodiment, the communication interface (240) may include at least one of a wired communication module or a wireless communication module.

[0066] The wired communication module may include various wired communication modules such as a Local Area Network (LAN) module, a Wide Area Network (WAN) module, or a Value Added Network (VAN) module, as well as various cable communication modules such as USB (Universal Serial Bus), HDMI (High Definition Multimedia Interface), DVI (Digital Visual Interface), RS-232 (recommended standard 232), power line communication, or POTS (plain old telephone service).

[0067] The wireless communication module may include a wireless communication module that supports at least one of a wireless communication method including WiBro (Wireless broadband), GSM (global System for Mobile Communication), CDMA (Code Division Multiple Access), WCDMA (Wideband Code Division Multiple Access), UMTS (universal mobile telecommunications system), TDMA (Time Division Multiple Access), LTE (Long Term Evolution), 4G, 5G, 6G, Bluetooth™, or Wi-Fi (Wireless-Fidelity).

[0068] In one embodiment, the processor (230) can obtain information from an external electronic device or server or provide information through a communication interface (240).

[0069] A processor (120) according to one embodiment can perform the operation of an electronic device (100) described below through the drawings. Specifically, the processor (120) can determine the possibility of collaboration between companies with high accuracy by efficiently integrating and analyzing various corporate data within an industrial cluster through RAG-based analysis and the use of AI agents. In addition, the processor (120) can simultaneously secure data security and privacy during the collaboration exploration process while maintaining the trust of the data provider companies through an anonymous-based intermediary protocol.

[0070] FIG. 3 is a diagram showing the system architecture and data processing flow of the electronic device (200) of the present invention.

[0071] Referring to FIG. 3, through a separated hierarchical structure in the system architecture, the electronic device (200) can reliably execute a series of processes including receiving a query, interpreting the semantics, searching for external context, generating a response, and performing subsequent tasks. Specifically, the system architecture connects the entire process from receiving a query, authentication, workflow selection, agent execution, tool invocation, context search, LLM generation, result streaming, and log storage via a standardized path, thereby reliably performing collaboration candidate discovery and matching judgment in an industrial cluster environment.

[0072] In FIG. 3, QUERY may refer to input submitted by a user, including natural language, parameters, and files. OUTPUT may refer to output in which the electronic device (200) returns a processing result to the user.

[0073] The Frontend Layer (310) may be a set of frontend components that display a user interface, maintain a session, and visualize a real-time status. The Frontend Layer (310) may include a UI, Chat Interface, Workflow Selection, and Real-time Status Display module and may pass a QUERY to an API path.

[0074] The API Gateway (320) may be an edge layer that accepts external requests and securely routes them to internal microservices. The API Gateway (320) may include a Load Balancer and an API Gateway module and may perform traffic distribution, rate limiting, and path policy application in the pre-authentication stage. The Real-time Layer (330) may be a streaming layer that provides WebSocket or SSE communication. The Real-time Layer (330) may transmit intermediate progress and messages of long-running tasks to the Frontend Layer (310) in real time.

[0075] The Core API Service (340) may be a core application layer composed of a FastAPI App, User Management, and Authentication Service. The Core API Service (340) may receive a QUERY delivered from the API Gateway (320), perform user authentication, normalize it into a standard request schema, and deliver it to the Agent Core Engine (360). The Core API Service (340) may check user permissions, organizational policies, and usage limits to initiate domain workflows within the allowed range.

[0076] Workflow Management (350) may be a workflow management layer composed of a Workflow Manager and a Template Store. Workflow Management (350) may provide the Agent Core Engine (360) with an execution specification that selects a standard template corresponding to the query intent and defines the steps. The Template Store may ensure reproducibility by storing domain-specific policies, preprocessing hooks, and postprocessing hooks along with versions.

[0077] The Agent Core Engine (360) may be an agent execution-centric layer composed of an Agent Core Controller, a State Tracking System, and a LangGraph Workflow Engine. The Agent Core Engine (360) can graph dependencies between stages, track state transitions, and perform retries, backoffs, and rollbacks in case of failure. The Agent Core Engine (360) can call Tool Management (370) at each stage to select and execute the necessary tools.

[0078] Tool Management (370) may be a tool orchestration layer composed of an MCP Server, Tool Execute, and Tool Registry. Tool Management (370) can select a suitable tool and manage the call lifecycle by referring to the schema, authorization tag, and call limit of the registered tool. Tool Management (370) can normalize the result to a standard output schema and return it to the Agent Core Engine (360).

[0079] Tools (380) may be a set of execution tools including a Tool Library, Internal APIs, External APIs, and a RAG Engine (Document Search). The RAG Engine of Tools (380) may be a search module responsible for document search and re-ranking, Internal APIs may perform internal system integration, and External APIs may perform external data integration. Tools (380) may comply with a call policy and return results and source metadata together.

[0080] The Data Layer (390) may be a data storage layer composed of a Primary Database and a Vector Database. The Primary Database may be a relational or document-based storage that stores Matching Data, Workflow Data, and Agent Memory. The Vector Database may include Private Collections, Public Collections, and Agent Memory, store embedding vectors, and perform similarity searches. The Data Layer (390) may ensure traceability by maintaining audit logs, access control, and version history.

[0081] Specifically, the Matching Data of the Data Layer (390) may be a core data layer for the electronic device to permanently store the entire process from query analysis, vector search, consistency or reliability evaluation, LLM generation, agent branching, anonymous brokerage, and security channel operation based on state and evidence. For example, the electronic device (200) can ensure explainability, real-time monitoring, and audit traceability by using the Matching Data to record candidate company similarity scores, evaluation metrics, weights, reordering rankings, collaboration potential scores, branching decisions, approval logs, and document exchange history in a consistent schema. For example, the electronic device (200) can automatically adjust ranking weights, reference matching scores, and policy rules by using success or rejection feedback accumulated in the Matching Data as a relearning signal.

[0082] LLM APIs can serve as call interfaces for large-scale language models. LLM APIs can perform context injection, summarization, inference, and format conversion, and support the generation of evidence-based responses. LLM APIs can apply system prompts, domain directives, and forbidden word rules to ensure response sensitivity and policy compliance.

[0083] In the data collection path, Contextual Data can refer to internal documents, wikis, logs, policy documents, and publicly available external materials. Data Pipelines are pipelines implemented using tools such as Databricks and Airflow that can perform collection, cleaning, and standardization. The Data De-identification module can generate non-identifiable data by masking or deleting sensitive identifiers. The Embedding Model can transform documents and records into high-dimensional vectors using embedding models such as OpenAI and Gemini. The Vector Database consists of tools such as Qdrant, Milvus, and Vespa, capable of handling embedding loading and Top-K nearest neighbor search.

[0084] Specifically, the electronic device (200) can enable the separate storage of private / public collections and policy-based access control by generating an anonymous ID through a data de-identification module and assigning standard metadata such as a non-identification status flag, risk score, industry code, authentication code, and region code. For example, the data de-identification module may be a core preprocessing component that detects direct identifiers such as company name, contact information, detailed address, and contract number, and quasi-identifiers such as rare equipment combinations and detailed coordinates from context data, and performs essential security cleansing before embedding and vector indexing by applying deletion, masking, generalization, and pseudonymization. By using the data de-identification module as a gateway for the data path, the electronic device (200) can simultaneously achieve anonymous matching, secure RAG search, and protection of personal information and corporate secrets.

[0085] In one embodiment, the Frontend Layer (310) may receive a QUERY and forward it to the API Gateway (320). The Core API Service (340) forwards the authenticated request to Workflow Management (350) and the Agent Core Engine (360), and the Agent Core Engine (360) may call the RAG Engine of Tool Management (370) and Tools (380) to retrieve relevant context from the Vector Database. The Agent Core Engine (360) may forward the retrieved context and query to LLM APIs to generate a list of collaboration candidates, a comparison table, and a suitability assessment report. The Core API Service (340) may convert the generated results into natural language, JSON, or visualization formats, stream them through the Real-time Layer (330), and display them on the Frontend Layer (310).

[0086] In terms of security and operation, the Core API Service (340) verifies authentication tokens and authorization tags, and the Tool Management (370) can apply rate limits and call limits when making external calls. The Data Layer (390) can store the mapping between anonymous IDs and real-name IDs separately and strictly control access. The Agent Core Engine (360) can record status and logs in the Workflow Data of the Primary Database to enable analysis of the cause of failure and reproduction.

[0087] In terms of scalability, each layer can be modularized to enable independent deployment, and functionality can be expanded without interruption by registering new tools in the Tool Registry. Vector Database can minimize latency even with large-scale data by adopting partitioning and indexing methods such as HNSW and IVF. The Real-time Layer (330) supports multiple subscriber streams so that users and operators can share the same progress.

[0088] FIG. 4 is a flowchart showing the operation of the electronic device (200) of the present invention.

[0089] Referring to FIG. 4, the electronic device (200) of the present invention analyzes input data including a user's query (operation 410), retrieves related data corresponding to the input data from a corporate information vector index in which at least one of the de-identified corporate technical information, certification status, and equipment list is vectorized and stored (operation 420), generates answer data based on the related data to provide an answer to the user's query (operation 430), and performs at least one subsequent task based on the context of the query and the answer using an AI agent (operation 440).

[0090] According to one example, in operation 410, the electronic device (200) can analyze input data including a user's query.

[0091] Here, the input data may be natural language queries, parameters, metadata, and session information transmitted by the user. The queries may include various purposes, such as searching for collaboration candidates, requesting facility sharing, verifying authentication match, and evaluating technical conformity.

[0092] The electronic device (200) may include a morphological analysis module, an entity extraction module, and an intent classification module to analyze input data. The morphological analysis module can identify meaningful words in a query by separating a sentence into tokens. The entity extraction module can extract entities directly related to industrial collaboration, such as facility names, technology names, region names, and certification names. The intent classification module can classify user purposes, such as collaboration requests, comparative analysis requests, or mediation requests, by combining session metadata and query configuration.

[0093] The electronic device (200) can standardize terms included in a query by referring to a predefined industry-specific knowledge base and domain dictionary during the input data analysis process. For example, if the term "medical robot" has the same meaning as "robotic equipment in the medical device field," the electronic device (200) can unify this into a standardized attribute value. The electronic device (200) can also maintain consistency in meaning by applying a multilingual processing module even when the user uses a mixed language (e.g., Korean and English) within the same query.

[0094] The electronic device (200) can remove unnecessary noise elements included in the input data. The unnecessary noise may be particles, conjunctions, punctuation marks, or common words with low semantic contribution. The electronic device (200) can improve the quality of the sentence and improve the accuracy of the analysis results by applying a normalization module to remove the noise.

[0095] The electronic device (200) can convert the result of analyzing input data into a structured form. The result converted into a structured form may include preprocessed data for generating a query vector, filter parameters for searching, and classification tags for determining the possibility of collaboration. For example, a query such as "Find a collaboration partner in Gyeonggi-do that has a robot arm facility and ISO certification" can be converted into structured data such as "Industry=Robot Arm, Certification=ISO, Region=Gyeonggi-do, Purpose=Search for Collaboration Partner".

[0096] The electronic device (200) can map key entities included in the query to attribute values ​​to generate structured results. The attribute values ​​may be multidimensional data such as industry sector, technology category, certification level, facility size, and geographical classification. The electronic device (200) can set filter parameters in a subsequent search step based on the attribute values.

[0097] The electronic device (200) can reflect session-based context during the input data analysis process. For example, if a user specifies "Gyeonggi-do region" in a previous conversation and then requests "find ISO-certified companies" in a new query, the electronic device (200) can recognize the session context and expand the query intent to "search for ISO-certified companies within the Gyeonggi-do region." Through session-based query expansion, the electronic device (200) can perform consistent analysis even in a continuous conversation situation.

[0098] The electronic device (200) can output the analyzed results along with a confidence score. The confidence score can be calculated based on query parsing accuracy, entity matching rate, and intent classification probability value. If the confidence score is below a threshold value, the electronic device (200) may request additional confirmation from the user or suggest a candidate for a corrected query.

[0099] The electronic device (200) can transmit the analysis results to the query vector conversion module and the search module. By analyzing the input data in multiple stages, the electronic device (200) can clearly define and classify the purpose of the query.

[0100] According to one example, in operation 420, the electronic device (200) can retrieve related data corresponding to the input data from a corporate information vector index in which at least one of the de-identified corporate technical information, certification status, and equipment list is vectorized and stored.

[0101] Here, the corporate information vector index may be a database that stores anonymized corporate data, such as technical information, certification status, facility lists, production capacity, and quality indicators, transformed into high-dimensional vectors through an embedding model. The vector index can be classified into private and public collections. The private collection may consist of data provided internally by a specific company. The public collection may consist of public data or externally disclosed materials.

[0102] The electronic device (200) can perform nearest neighbor search in the same vector space as the corporate information vector index by generating a query vector after preprocessing input data. The electronic device (200) can apply various metrics, such as cosine similarity, dot product, and distance-based calculation, to calculate the similarity between the query vector and the corporate data vector. The electronic device (200) can limit search results by deriving K corporate vectors closest to the query vector as candidates or by selecting vectors whose similarity score is above a threshold.

[0103] The electronic device (200) may receive metadata associated with the searched results. The metadata may include various items such as corporate anonymous ID, data source, time of recording, industry code, certification tag, facility size, quality indicator, supply chain linkage information, maintenance history, and safety compliance. The electronic device (200) may use the metadata to evaluate the consistency and reliability of the search results and, if necessary, readjust the ranking based on weights.

[0104] The electronic device (200) can apply filter parameters specified by the user during the search process. The filter parameters may consist of an industry sector, technology category, certification level, facility size, geographical location, etc., and the electronic device (200) can improve the accuracy of the results by excluding corporate data that does not satisfy the filter conditions. In addition, the electronic device (200) can ensure the timeliness of the data by prioritizing the latest data through time-series filtering and reducing the weight of past records.

[0105] The electronic device (200) can search private collections and public collections in parallel and integrate the respective results to generate a list of corporate information. Parallel searching improves the speed of data access, and during the result integration process, duplicate data can be removed or data from multiple sources regarding the same company can be merged. The list of corporate information may include multiple collaboration candidates that correspond semantically to the input data, and each company can be identified by an anonymous ID to minimize the risk of personal information leakage.

[0106] The electronic device (200) may score source reliability, industrial conformity, technical similarity, and certification match to evaluate the reliability of the retrieved data, and provide a final sorted list by weighting and summing. If the search result is below the standard reliability score, the electronic device (200) may exclude the result or classify it into a separate 'pending list' to wait for the user's selection.

[0107] The electronic device (200) can reflect a Retrieval-Augmented Generation (RAG) structure during the search process. The RAG structure serves to configure an extended context so that the searched related data can be directly passed to the LLM-based answer generation stage. Therefore, the search result in operation 420 can go beyond simple data retrieval and serve as a basis for subsequent answer data generation and subsequent task execution.

[0108] In this way, in operation 420, the electronic device (200) searches for data that is semantically most similar to the input query in de-identified large-scale enterprise data and provides metadata included in the searched data, thereby enabling more accurate and reliable collaborative analysis and subsequent task execution in subsequent steps.

[0109] According to one example, in operation 430, the electronic device (200) can provide an answer to the user's query by generating answer data based on the relevant data.

[0110] Here, the response data is a result generated by combining searched company information and user queries, and may include results in various formats such as a list of collaboration candidates, a comparative analysis table, a suitability assessment report, a technical summary, and whether certifications match.

[0111] The electronic device (200) can construct an extended context by integrating the query and the retrieved related data to generate answer data. The extended context may be a set of data that combines multiple attributes, such as core keywords, industry sectors, certification conditions, and facility conditions extracted from the query, and attributes such as technical descriptions, facility specifications, certification records, quality indicators, supply chain linkage information, geographic location information, timestamps, and source reliability collected from the retrieved corporate data. The electronic device (200) can generate more precise and richer answer data by using the extended context in this way as an input for an LLM API call.

[0112] The electronic device (200) can generate natural language responses by injecting a query and an extended context through the LLM API. The generated response may go beyond simple question-and-answer and include analysis results such as a list of collaboration candidates, comparative analysis results, and a suitability evaluation report, and, if necessary, may include explanatory responses explaining why a specific company is suitable for collaboration or its technical strengths.

[0113] The electronic device (200) can convert response data into various formats suitable for the user environment. For example, the response data can be converted into structured data (JSON format) to enable linked processing with other systems, or converted into visualization data format to be displayed as visual elements such as graphs, tables, and diagrams. Additionally, responses in natural language format can be provided in a user-friendly manner in a UI environment.

[0114] The electronic device (200) can correct the reliability of the data retrieved during the response data generation process. Reliability correction can be performed by applying weights based on indicators such as industry suitability, technical similarity, certification reliability, recency, and source reliability, and the correction results can be reflected in the alignment and priority setting of the response data. For example, data with low recency or low source reliability may be placed in a lower priority in the response results.

[0115] In one embodiment, the electronic device (200) can generate response data including at least one of a collaboration candidate list, a comparison analysis table, and a suitability evaluation report by configuring an extension context and calling an LLM API, and can convert the generated response data into at least one of a natural language format, a structured data format, or a visualization data format to provide a result optimized for the user environment.

[0116] The electronic device (200) can utilize the generated answer data as input for performing subsequent tasks. For example, the generated list of collaboration candidates can be passed to a corporate matching procedure in a subsequent stage, and the comparison analysis table and suitability evaluation report can be used to determine collaboration possibilities or user review processes. Thus, operation 430 is not merely a stage for generating simple question-answers, but can serve as a key link for the execution of subsequent AI agent-based tasks.

[0117] As a result, the electronic device (200) can provide accurate and explainable results that meet the purpose of the user's query by configuring an extended context that considers the consistency and reliability of the data retrieved in operation 430 and by generating and converting answer data of various formats based on the LLM API.

[0118] According to one example, in operation 440, the electronic device (200) can use an AI agent to perform at least one subsequent task based on the context of the query and the answer.

[0119] The electronic device (200) can perform various follow-up tasks based on the context of the query and the answer using an AI agent. For example, the follow-up tasks may include various tasks such as business operations, data analysis, policy compliance, risk management, or determining the possibility of collaboration with other companies of companies included within a specific industry cluster.

[0120] The electronic device (200) can perform various analysis tasks using an AI agent, such as automatically generating industry trend reports, evaluating compliance with the latest certification regulations, analyzing supply chain risks, and matching with companies capable of collaboration. The electronic device (200) can configure a real-time monitoring dashboard based on additional user requests and run a specific technology trend prediction model based on search results. The electronic device (200) can update user-customized training data or identify recurring query patterns to suggest an automated set of recommended queries.

[0121] The electronic device (200) can automate subsequent tasks across the industry, such as proposing predictive maintenance schedules, deriving energy efficiency optimization methods, checking quality management processes, calculating ESG-related indicators, and evaluating the possibility of inter-company collaboration, through an AI agent. The electronic device (200) can convert the generated results into a natural language format, a structured data format, or a visualization data format and provide them to the user. Additionally, the electronic device (200) can utilize the results for future analysis and policy formulation by linking them with log data and managing them for long-term accumulation.

[0122] In one embodiment, the electronic device (200) may use an AI agent to determine the possibility of collaboration between companies within an industrial cluster based on the context of the query and the answer. For example, the at least one subsequent task may include determining the possibility of collaboration between a first company within a first industrial cluster and at least one other company within the first industrial cluster based on at least one of the input data, the relevant data, and the answer data.

[0123] FIG. 5 is a conceptual diagram showing the operation of the electronic device (200) of the present invention matching companies based on the possibility of collaboration, and FIG. 6 is a flowchart showing the operation of the electronic device (200) of the present invention matching companies based on the possibility of collaboration.

[0124] Referring to FIGS. 5 and 6, the electronic device (200) can generate a list of company information (operation 610), extract at least one company to be collaborated with by evaluating the consistency between multiple companies (operation 620), calculate the possibility of collaboration for the company to be collaborated with (operation 630), and determine whether a match exists by comparing the possibility of collaboration with a reference matching score (operation 640).

[0125] In operation 610, the electronic device (200) can generate a list of corporate information. The list of corporate information may be a data set containing vector values ​​of retrieved non-identifiable corporate data and metadata associated therewith. The electronic device (200) can merge data retrieved from private collections and public collections and sort the list according to predefined filter parameters such as industry sector, technology category, certification level, and facility size. The electronic device (200) can manage the companies included in the list based on anonymous identifiers so that personal and sensitive information is not exposed.

[0126] In operation 620, the electronic device (200) can evaluate the consistency between multiple companies and extract at least one company as a candidate for collaboration. For example, the electronic device (200) can evaluate the consistency between companies by considering at least one of the following: suitability for the industrial sector, technical similarity, certification consistency, degree of reflection of the latest data, and source reliability. The electronic device (200) can compare each company's data with reference company data and calculate a weighted consistency score to extract the top company as a candidate for collaboration. By integrating multiple evaluation indicators to derive a final candidate, the electronic device (200) can identify companies with high actual collaboration suitability, rather than merely calculating similarity.

[0127] In operation 630, the electronic device (200) can calculate the possibility of collaboration for a target company. The possibility of collaboration may be a collaboration matching score calculated by combining multiple indicators, such as a technology compatibility score, a certification conformity score, a facility utilization possibility score, a supply chain linkage indicator, and a quality management indicator. The electronic device (200) can perform a reliability correction procedure during the evaluation process to correct the results based on the data's recency or source reliability. By quantifying the results of the possibility of collaboration, the electronic device (200) can quantitatively compare the possibility of collaboration between candidate companies.

[0128] In operation 640, the electronic device (200) can determine whether a match is made by comparing the possibility of collaboration with a reference matching score. The reference matching score may be a value set through a predefined threshold or a dynamic learning model. The electronic device (200) can determine that a match has been made if the possibility of collaboration is greater than or equal to the reference matching score, and can perform alternative company search or recommend a new partner if the possibility of collaboration is less than the reference matching score. The electronic device (200) can notify the user in real time whether a match is made and can link to anonymous connection requests, sending collaboration proposals, or generating reports.

[0129] FIG. 7 is a diagram showing a mediation protocol in which the electronic device (200) of the present invention mediates between a plurality of matched companies.

[0130] Referring to FIG. 7, the electronic device (200) can create a secure collaboration channel between multiple companies (e.g., Company A and Company B) based on a cloud environment and securely mediate the entire collaboration process through the secure collaboration channel.

[0131] The electronic device (200) can use cloud infrastructure to provide a secure, anonymous-based intermediary path instead of a direct connection between companies. The electronic device (200) can perform initial communication without exposing the identities and sensitive information of Company A and Company B using an anonymous identifier and a temporary session key. Additionally, the electronic device (200) can apply an encryption module to manage all transmitted data so that it is protected from unauthorized access and tampering.

[0132] The electronic device (200) can transmit key data necessary for the execution of collaboration, such as a schedule draft, a task list, and a data exchange policy, through a secure collaboration channel. For example, multiple matched companies (e.g., Company A and Company B) can exchange agreements regarding their respective collaboration roles, execution schedules, and methods of data sharing through the secure collaboration channel. Additionally, the electronic device (200) can securely manage transmitted and received data by tracking and recording the transmitted data in real time.

[0133] The electronic device (200) can automatically store document exchange history and approval logs generated in a secure collaboration channel. By recording logs such as collaboration proposals, contract drafts, material updates, and approval status along with timestamps, the electronic device (200) can be utilized for future collaboration tracking and audit verification. Additionally, the electronic device (200) can manage the stored log data as integrated matching data by linking it with search grounds, technical tags, and inter-company complementarity analysis results generated during the collaboration matching process.

[0134] The electronic device (200) can combine the document exchange history and approval logs generated during the process of facilitating inter-company collaboration with the relevant technology tags or corporate complementarity analysis results derived from the previous search process. Therefore, the electronic device (200) can go beyond simple communication mediation and record the matching basis and the progress of collaboration together, thereby improving the accuracy of the collaboration recommendation model for similar queries in the future.

[0135] As such, the electronic device (200) of the present disclosure and the collaborative company matching method using the same can efficiently integrate and analyze various company data within an industrial cluster to determine the possibility of collaboration between companies with high accuracy. Since the electronic device (200) of the present disclosure and the collaborative company matching method using the same can reflect multidimensional data attributes through RAG-based analysis and the use of AI agents, it is possible to evaluate the possibility of collaboration in a sophisticated manner that was lacking in conventional technology. Therefore, the electronic device (200) of the present disclosure and the collaborative company matching method using the same can significantly improve the efficiency of the process of discovering collaboration between companies.

[0136] In addition, the electronic device (200) of the present disclosure and the collaborative company matching method using the same allow sensitive data to be safely utilized because the company information is processed in a de-identified state. The electronic device (200) of the present disclosure and the collaborative company matching method using the same can simultaneously secure data security and privacy during the collaboration search process while maintaining the trust of the data provider company through an anonymous-based intermediary protocol. Therefore, companies within an industrial cluster can review the possibility of collaboration with other companies without fear of data leakage.

[0137] In addition, the electronic device (200) of the present disclosure and the collaborative company matching method using the same can provide the result of determining the possibility of collaboration as a highly reliable indicator, thereby increasing the likelihood of leading to actual collaboration. Since the electronic device (200) of the present disclosure and the collaborative company matching method using the same can effectively match candidate companies for collaboration based on the result of evaluating the possibility of collaboration, new partnership opportunities within an industrial cluster can be secured more quickly and stably.

[0138] However, as this has been explained above, a redundant explanation thereof will be omitted.

[0139] Using the embodiments of the present invention described above, those skilled in the art will be able to easily make various changes and modifications within the scope of the essential characteristics of the present invention. The content of each claim of the patent claims may be combined with other claims that are not related by reference within the scope of what can be understood from this specification.

Claims

Claim 1 An electronic device that generates an answer to a user's query based on RAG (Retrieval-Augmented Generation) and performs a subsequent task using an AI agent, comprising: a memory that stores at least one instruction; and includes at least one processor that executes the at least one command, wherein the at least one processor analyzes input data including a user query, retrieves related data corresponding to the input data from a corporate information vector index in which at least one of anonymized corporate technical information, certification status, and facility list is vectorized and stored, and provides an answer to the user query by generating answer data based on the related data, and performs at least one subsequent task based on the context of the query and the answer using an AI agent, wherein the at least one subsequent task includes determining the possibility of collaboration between a first company within a first industrial cluster and at least one other company within the first industrial cluster based on at least one of the input data, the related data, and the answer data, wherein the possibility of collaboration is determined based on at least one of a technology compatibility score, a certification conformity score, a facility utilization possibility score, a supply chain linkage indicator, and a quality management indicator, and wherein the at least one processor sequentially performs the derivation of collaboration matching candidates, comparative analysis, and an anonymous connection request when the possibility of collaboration is greater than or equal to a reference matching score, and wherein the possibility of collaboration is greater than the reference matching score If less than, an electronic device that performs alternative industry cluster exploration or new partner recommendation. Claim 2 An electronic device according to claim 1, wherein the at least one processor provides an anonymous-based mediation protocol that mediates collaboration between the first company and the at least one other company matched with the first company. Claim 3 An electronic device according to claim 1, wherein at least one processor tokenizes a morphological analysis module through the input data, extracts entities and attributes for the query included in the input data, and determines the intent and subject of the query based on session metadata. Claim 4 An electronic device according to claim 1, wherein at least one processor converts the query included in the input data into a high-dimensional query vector using an embedding model, and sets filter parameters including at least one of an industry field, a technology category, a certification level, and a facility size associated with the query vector. Claim 5 An electronic device according to claim 1, wherein at least one processor performs a nearest neighbor search for a non-identifying corporate data vector mapped to the same vector space as the query vector and extracts at least one corporate anonymous identifier based on a similarity score. Claim 6 An electronic device according to claim 1, wherein at least one processor searches a private collection and a public collection in parallel and sorts the private collection and the public collection in a preset sorting manner to generate a corporate information list. Claim 7 An electronic device according to claim 1, wherein the relevant data comprises at least one of a de-identified enterprise technical description, equipment specification information, certification history information, production capacity indicators, quality control indicators, process standard documents, safety compliance records, maintenance history, supply chain linkage information, geographic location information, timestamps, and source metadata. Claim 8 An electronic device according to claim 1, wherein at least one processor calculates at least one of industrial suitability, technical similarity, certification reliability, timeliness, and source reliability for the relevant data, and readjusts the result ranking of the relevant data based on the calculation result. Claim 9 An electronic device according to claim 1, wherein at least one processor combines the relevant data and the query to form an extended context and, by calling an LLM API, generates the answer data including at least one of a collaboration candidate list, a comparative analysis table, and a suitability evaluation report. Claim 10 An electronic device according to claim 1, wherein at least one processor converts the answer data into at least one of a natural language format, a structured data (JSON) format, and a visualization data format, and transmits the converted answer data in a format optimized for the user environment. Claim 11 delete Claim 12 An electronic device according to claim 1, wherein at least one processor transmits an initial inter-enterprise collaboration request matched based on an anonymous ID, exchanges real-name data only when approval for the collaboration request is made between enterprises, and provides an execution status and result log of at least one subsequent task. Claim 13 An electronic device according to claim 1, wherein at least one processor creates a secure collaboration channel between the first enterprise and the first enterprise, transmits a schedule draft, a task list, and a data exchange policy through the secure collaboration channel, and stores the document exchange history and approval log of the secure collaboration channel in conjunction with matching data including at least one of a search basis, a technical tag, and an inter-enterprise complementarity analysis generated during the collaboration matching process. Claim 14 A method for generating an answer to a user's query based on RAG (Retrieval-Augmented Generation) and determining the possibility of inter-company collaboration using an AI agent, comprising: a step of analyzing input data including a user's query; a step of retrievaling related data corresponding to said input data from a corporate information vector index in which at least one of anonymized corporate technical information, certification status, and equipment list is vectorized and stored; and a step of providing an answer to said user's query by generating answer data based on said related data. A method for matching collaborative companies, comprising: a step of performing at least one subsequent task based on the context of the query and the answer using an AI agent; wherein the at least one subsequent task includes determining the possibility of collaboration between a first company within a first industrial cluster and at least one other company within the first industrial cluster based on at least one of the input data, the relevant data, and the answer data, wherein the possibility of collaboration is determined based on at least one of a technology compatibility score, a certification conformity score, a facility utilization possibility score, a supply chain linkage indicator, and a quality management indicator; and wherein the step of performing the at least one subsequent task includes: a step of sequentially performing a collaboration matching candidate derivation, a comparative analysis, and an anonymous connection request when the possibility of collaboration is greater than or equal to a reference matching score; and a step of performing an alternative industrial cluster search or a new partner recommendation when the possibility of collaboration is less than the reference matching score. Claim 15 A method for matching collaborative companies, wherein the step of performing at least one subsequent task comprises: providing an anonymous-based mediation protocol that mediates collaboration between the first company and at least one other company matched with the first company. Claim 16 delete Claim 17 A method for matching collaborative companies according to claim 14, wherein the step of performing at least one subsequent task comprises: transmitting an initial collaboration request between companies matched based on an anonymous ID; exchanging real-name data only when approval for the collaboration request is made between companies; and providing an execution status and result log of at least one subsequent task. Claim 18 In claim 14, the step of performing at least one subsequent task comprises: creating a secure collaboration channel between the first enterprise and the first enterprise; transmitting a schedule draft, a task list, and a data exchange policy through the secure collaboration channel; and storing the document exchange history and approval log of the secure collaboration channel in conjunction with matching data including at least one of a search basis, a technical tag, and an inter-enterprise complementarity analysis generated during the collaboration matching process; a collaborative enterprise matching method.

Images