KR-20260065603-A - Milestone-linked knowledge collaboration mediation system and method comprising AI-based problem standardization engine and hierarchical authority separation-based data isolation logic

Abstract

The present invention relates to a knowledge collaboration brokerage system that normalizes R&D challenges in natural language form into standard technical objects through AI, automatically concludes Non-Disclosure Agreements (NDAs) and records blockchain integrity prior to the commencement of collaboration, and provides a zero-trust security environment that logically isolates the scope of data access between a corporate master and individual members. In addition, the present invention is characterized by enhancing corporate security and maximizing the transparency of expert matching and settlement by recording integrity proof information of the transaction on a distributed ledger along with the transfer of virtual assets or fiat currency at the time of achieving a mutually agreed visual milestone.

Inventors

• 이동일

Assignees

• 이동일

Dates

Publication Date

20260508

Application Date

20260421

Claims (4)

1. A knowledge collaboration mediation method performed by a computer server, comprising: (a) a step of receiving problem data in the form of unstructured natural language from a user terminal and converting it into a structured project object containing technical parameters and performance indicators (KPIs) through a Large Language Model (LLM)-based engine; (b) a step of matching a client terminal and a problem solver terminal based on the converted project object, receiving an electronic signature for an automatically generated Non-Disclosure Agreement (NDA) object, and recording the hash value of the agreement in a distributed ledger; (c) a step of separating corporate Master authority and individual Member authority based on user identification information, allocating a virtual collaboration space in which the data access scope for each authority is isolated, wherein the Master authority holder is restricted from accessing practical work data and is granted access only to the final deliverable; and (d) a step of disbursing funds deposited in the server to the problem solver terminal and recording the hash value of the technical data related to the milestone in the distributed ledger upon the occurrence of an approval event for the completion of a Milestone set within the virtual collaboration space.
2. A knowledge collaboration brokerage method according to claim 1, wherein step (c) logically separates the 'member context' when participating in a project affiliated with a corporation and the 'independent context' when engaging in individual expert activities for the same personal identifier (ID), and integrates and manages only the reputation score among the activity history generated in each context.
3. A knowledge collaboration brokerage method according to claim 1, wherein, at the time of funding payment in step (d), the server generates transaction identification information (Transaction Hash) for proving the time of creation of technical data generated during the collaboration process, the hash value of the technical data, and a timestamp, in conjunction with a milestone approval event, and records them in the distributed ledger.
4. A knowledge collaboration brokerage method according to claim 1, further comprising: a step of extracting a real-time hash value from data recorded in a storage when a data call request is received from a user terminal; a step of verifying integrity by comparing whether the real-time hash value matches the original hash value previously recorded in the distributed ledger; and a step of decrypting and providing only the data for which the integrity verification is completed.

Description

Milestone-linked knowledge collaboration mediation system and method comprising AI-based problem standardization engine and hierarchical authority separation-based data isolation logic Milestone-linked knowledge collaboration mediation system and method comprising AI-based problem standardization engine and hierarchical authority separation-based data isolation logic The present invention relates to a brokerage system combining information processing using artificial intelligence and settlement technology based on a distributed ledger, and more specifically, to a knowledge collaboration brokerage system and method that standardizes unstructured difficult problems using an AI engine and settles funds according to milestones in a Zero-Trust security environment. Existing expert matching platforms face limitations, such as the difficulty in finding suitable experts due to clients' vague requirements, and security vulnerabilities that allow for the leakage of a company's core technological ideas during the collaboration process. Additionally, a lack of transparency in the payment process based on work performance leads to frequent disputes between clients and problem solvers. FIG. 1 is an overall system configuration diagram according to an embodiment of the present invention. FIG. 2 is a diagram showing hierarchical authority isolation and data security logic according to an embodiment of the present invention. FIG. 3 is a flowchart illustrating an AI-based problem refinement process according to an embodiment of the present invention. FIG. 4 is a flowchart illustrating an integrated collaboration and settlement workflow according to an embodiment of the present invention. Detailed description of FIG. 1: 100 (User terminal unit) is an end-user terminal device comprising a client (corporate master) terminal that performs a task request and a problem solver (individual member) terminal that performs a task solution, and is connected to the SolveNet integrated server (200) via a wired or wireless network. 110 (Corporate Master) refers to an account or terminal that requests a technical task requiring a solution and receives and approves the final deliverable. 120 (Individual Member) is an expert terminal that is matched to the requested task, derives a technical solution, and uploads deliverables for each milestone. 200 (SolveNet integrated server) is a central control system that controls the logic of the entire system and manages the matching, contract, security, and settlement processes in an integrated manner. 210 (AI problem refinement engine) is a core module that collects unstructured requirements in the form of natural language input from the user and extracts technical field mapping and quantitative performance indicators (KPIs) through LLM-based semantic analysis. 220 (Automatic NDA and Contract Generation Module) automatically generates legally binding NDAs upon successful matching, manages electronic signatures, and extracts and transmits the hash value of the generated contract. 230 (Authority Control and Security Module) manages system access rights by user type, and specifically logically isolates the scope of practical data access between the corporate master and individual members, and performs encryption of stored data. 300 (Distributed Ledger System) guarantees data integrity by recording the hash value of concluded contracts and transaction timestamps on blockchain nodes. 400 (Integrated Payment Gateway) performs escrow functions for fiat currency and virtual assets, and executes payment in response to milestone approval events. Detailed description of FIG. 2. The present invention performs hierarchical authority isolation based on a Zero-Trust security model. All data stored in 250 (Zero-Trust-based encrypted storage) remains encrypted by default. 120 (Individual Member) has free access and editing rights to the work data (draft) they are currently working on, but 110 (Corporate Master) is logically blocked (Access Blocked) from accessing the work data by a 'data isolation barrier'. The access rights of the Corporate Master are structured such that decryption and access to the data are permitted only when the milestone is completed and the final result is approved. Detailed description of Fig. 3: When a user inputs a problem in ambiguous natural language at 310 (Step 1), an LLM-based AI engine analyzes the context at 320 (Step 2) to identify the intent. At 330 (Step 3), based on the analyzed intent, quantitatively measurable technical performance indicators (KPIs) and detailed parameters are derived, and based on this, at 340 (Step 4), a structured data object named 'Standard Project Technical Specification' is created and registered in a database. Detailed description of FIG. 4: After 410 (successful expert matching), an NDA contract is automatically generated by 220 (automatic NDA and contract generation module), and step 420 (automatic non-disclosure agreement generation) is completed thr