AUTONOMOUS INFORMATION LINE AND LLM VEHICLE DEFINITION GENERATION METHOD AND SYSTEM

Abstract

This invention relates to a method and system (1) for generating an autonomous knowledge graph and LLM tool definition between one or more servers (2) to which one or more clients (3) are connected, for use in software systems that generate request-response based communication logs, event logs or inter-service communication traces, including but not limited to microservice-based enterprise software systems, cloud-based service platforms, application programming interface-intensive systems, event-driven data processing systems, and large language model-supported automation and decision support systems.

Description

1 TARIFF Generating Autonomous Information Line and LLM Vehicle Identification METHOD AND SYSTEM This invention enables request-response based communication logs, event logs, or inter-service 5 In software systems that generate communication traces, including but not limited to microservices. in enterprise software systems, cloud-based service platforms, applications programming interface intensive systems, event-driven data processing systems, large language For use in model-assisted automation and decision support systems, one or more Autonomous information graph between one or more servers (2) to which multiple clients (3) are connected 10 and is related to the method and system of generating LLM vehicle definition (1). In traditional software-based systems, service definitions are part of application programming. Interface conventions, data entity relationships, and integration interfaces are mostly software-based. This is manually documented by developers or system operators. Similar 15 Similarly, tool definitions used in large language model-based integrations are also typically created manually. They are created and stored in a static form. In these systems, the operational process is more efficient. Since many people are involved, integration costs increase. Also, user In large language model integrations written manually by the large language model tool As definitions remain static, structures susceptible to human error also emerge during this process. 20 This concludes. Following system changes, the relevant documents and integration codes are provided. It needs to be constantly updated. The processes are largely dependent on user intervention. In these systems, documents quickly become outdated, which affects the actual system behavior. This can lead to inconsistencies between the documents. Current application programming. Interface documentation tools, such as OpenAPI or Swagger-like approaches, are primarily based on 25 This reflects the static schema definitions provided by the developer and does not represent the actual system. It cannot automatically discover the timely behavior of the service network or distributed system. Observability solutions analyze inter-service traffic topology and call relationships. It can track the distribution of data assets across the system, at the asset level. By extracting relationships and endpoint semantics, these can be used for a large language model. 30 It cannot generate tool definitions. In large language model tool frameworks, function calls are used. or vehicle usage schemes are often defined statically by developers, Manual updates are required following system changes. Therefore, the actual Over time, a mismatch develops between system behavior and vehicle definitions, thus 2 This increases integration costs and leads to the emergence of structures prone to human error. This is done. In addition, in distributed software systems, new endpoints are added, and existing endpoints are expanded. Changes in data payload structures at the points, the emergence of new data assets, services As a result of the evolution of data flow patterns or changes in system topology, manual The existing tool definitions and documentation quickly become outdated. This situation affects major language groups. model-based systems interacting with target systems in a secure, accurate, and up-to-date manner. This makes it difficult to access data assets from live system traffic. Therefore, at the edge... points, data flows between services, and the semantics of create, read, update, and delete. automatically extracts, creates an up-to-date information graph from, and adds it to the information graph. System 10 generates structured tool definitions for large language models based on... an integrated system that automatically updates these tool definitions by continuously monitoring their changes. A system and method is needed. One or more clients (3) are connected to or autonomous knowledge graph and LLM tool definition used between multiple servers (2) with the production method and system (1), manual documentation, manual tool production and By reducing the number of updates, microservices, application programming interface gateway and 15 Request-response records, event logs, service records obtained from target systems (4) such as databases Using call traces and similar traffic data, data assets, endpoints, and services are analyzed. automatic data flows and creation, reading, updating, and deletion relationships between devices extracting this information, creating an up-to-date information chart from it, this information Generating current and structured tool definitions for large language models from the graph, 20 Continuous monitoring of structural changes occurring in the system, and the definitions of the tools generated. Verification and secure updating of approved vehicle definitions. installation in the environment without service interruption and rollback of faulty versions It is intended to be accomplished. Generating autonomous knowledge diagrams and LLM vehicle definitions to achieve the invention's objective. The method and system (1) are formulated in the attached figures, from which; Figure 1. Management and system for generating autonomous knowledge graph and LLM vehicle definition. The parts in the figures are numbered, and their corresponding parts are shown below: 1. Method and system for generating autonomous knowledge graph and LLM vehicle definition. 2. One or more servers 3 3. One or more clients 4. Target systems 5. Traffic collection module 6. Traffic Separation Module 7. Asset and creation, reading, updating and deleting analysis module 5 8. Information graph repository 9. LLM Vehicle Identification Manufacturer 10. LLM work environment 11. Continuous monitoring module 12. Verification and rollback module 10 13. Network communication units 14. Function complement 15. Argument diagram 16. Turning structure 17. Natural language explanation 15 Method and system for generating autonomous information graph and LLM vehicle definition (1), one or multiple servers (2), one or more clients connected to one or more servers (2) (3), target systems (4), traffic collection module (5), traffic separation module (6), assets and Create, read, update and delete analysis module (7), information graph repository (8), LLM tool 20 definition manufacturer (9), LLM operating environment (10), continuous monitoring module (11), verification and feedback receiving module (12), network communication units (13) that provide data communication between modules, function complement (14), argument scheme (15), return structure (16) and natural language It consists of explanation (17). Management and system for generating autonomous information graph and LLM vehicle definition (1), one or one or more clients (3) are connected via network communication units (13) It runs on server (2). Method of generating autonomous knowledge graph and LLM vehicle definition. and system (1) includes multiple software components and each software component has a different or It can run on multiple servers (2). Autonomous knowledge graph and LLM tool definition 30 method and system of production (1) from multiple microservices, each implementing a different service Because of this, these software components consider load balancing, accessibility, and scalability. According to the rules, it is located on one or more servers (2). Each software This component exchanges data with other software components and, depending on the operation performed, also generates a response. 4 or multiple server (2) and one or more client (3) roles can be changed. In fact, the method and system for generating autonomous knowledge graph and LLM vehicle definition (1), third party software systems, adapters, connector interfaces, plugins, network hook components, or They are integrated and communicate through intermediary services. between one or more servers (2) to which one or more clients (3) are connected Autonomous information diagram and LLM vehicle definition generation method and system used (1) traffic collection module (5), microservice, application programming interface gateway and databases raw request-response traffic, event logs and service interaction from target systems (4) such as It is the unit that collects the data. This data collection process is carried out using the application programming interface network 10. gateway logs, reverse proxy logs, service network tools, application-level middleware logs, distributed monitoring data, message queue consumer listeners, event stream observation This can be done through components or similar resources. Where necessary, database-related transactions, query logs, change data capture records, or transaction data. It can also be included in the system via its tracks. Thus, the traffic collection module (5), 15 raw data representing the actual working behavior of the target systems (4) central or It logically aggregates into a unified observation layer. Traffic parsing module (6), By processing the raw records obtained by the traffic collection module (5), structured events It converts into records. During this conversion, the traffic parsing module (6), endpoint identifiers, HTTP methods or equivalent call types, path patterns, query 20 parameters, request payload schemas, response payload schemas, response structures, timing their stamps, correlation identifiers, trace identifiers, and persistent entity identifiers It extracts the persistent entity identifiers extracted by the traffic parsing module (6), response or recurring characters in request bodies such as id, uuid, key, slug that are suitable for distinguishing an entity. Statistical repeat analysis of unique fields such as, field pattern extraction, rule-based 25 This is determined using methods or learning methods. In an application form, The persistent entity identifiers generated by the traffic parsing module (6) are the field values. format pattern, frequency of repetition, consistency between sources, temporal stability, and distinctiveness It is determined by using the level. The traffic separation module (6) also has different By normalizing data from various sources, the system achieves a common event format. by transforming and thus enabling the simultaneous analysis of heterogeneous records. It makes it possible. Entity and creation, reading, updating and deleting analysis module (7), traffic parsing data entities, services, through structured records provided by module (6), semantic relationships between endpoints, the expansions of existence, and the processes of creation and reading, It identifies update and delete semantics. Entity and creation, read, update. and the deletion analysis module (7), persistent asset 5 extracted by the traffic parsing module (6). By using identifiers, we can define data entities and allow the same entity value to be used at different ends of the spectrum. at points, across different services, and throughout temporally related request-response sequences. By observing how things are seen, it determines the spread of entities. Thus, existence and creation, With the read, update and delete analysis module (7), it can be determined which data entity is in the system. created, read, modified, deleted or otherwise used by endpoints and other services and 10 It can be determined that it is transferred to endpoints. Entity and creation, reading, updating and The delete analysis module (7) analyzes the create, read, update and delete semantics only at the end. not through superficial matching based on dot name or HTTP method, but through consecutive matchings belonging to the same entity. or it is determined by comparing related request-response pairs. In this context, Response differential technique is applied, the first time a new entity identifier is seen 15 creation, returning an existing entity unchanged, reading, difference in field values The observed update means that the entity in question is no longer returned in subsequent related requests. Or, if it is inaccessible, it is interpreted as a deletion. Depending on the need, this Statistical methods in the analyses, rule-based inference mechanisms, embedding vector. Approaches or machine learning classifiers can be used. Entity and 20 Create, read, update and delete analysis module (7), extracted endpoint semantics, assigning confidence scores to entity relationships and parameter types, low confidence results It can redirect to user review. Autonomous knowledge graph and LLM vehicle definition generation method and system (1) knowledge graph 25 The data repository (8) is the unit in which the analyzed data is stored in the form of nodes and edges. Information In the graph store (8), nodes represent services, endpoints and data entities, Additional node types for schema structures, process patterns, or event types as needed. It can be created. Calls include reading, writing, creating, updating, deleting, and transferring. Edges that can be used in this type of configuration represent call relationships and data flows between nodes, 30 It represents the semantics of create, read, update, and delete, as well as entity deployments. On each node and edge, the data type, field set, confidence score, resource timestamp, and initial Attributes such as time of sighting, last sighting time, and source trace can be stored. Confidence scores reflect the certainty of the relevant inference and are above the determined threshold value. 6 The fields or parameters below are for additional validation or user review. can be marked. In an application format, the extracted parameter types, entity relationships, and Confidence scores for endpoint semantics, number of observations, source diversity, schema. consistency, the field being viewed consistently at different extremes, and repetitive use. It is calculated by weighted evaluation of patterns. Autonomous knowledge graph and LLM 5 In the method and system of generating vehicle definition (1), the information graph store (8) is only the system. Beyond being a structure in which the topology is stored, it also contains contextual relationships and data at the entity level. It is a queryable unit representing the flow of information. With the information graph store (8), a specific data which endpoints its existence is produced by, which services it is transferred between, which path-based resource 10, modified in steps or consumed by which endpoints This allows the system's behavior to be traced retrospectively through queries. It becomes explainable and the necessary contextual ground for vehicle definition production. is provided. While generating vehicle definitions from the information graph repository (8), call sequences The statistical usage frequency is measured and only when it exceeds a certain repetition threshold and the system When converting sequences of operations that form a consistent pattern into a tool, random or biased 15 Calls are filtered. LLM vehicle definition manufacturer (9), information sheet stored in (8) By using endpoint, entity, schema, and relational information, by large language models It generates usable, structured tool definitions. These tool definitions are specific to a particular purpose. a non-provider-specific and structured function call or tool call They are created as diagrams that can be interpreted in various formats. Each tool definition has at least one 20-digit number. The function includes the function descriptor, argument scheme, return structure, and natural language description. The LLM tool definition manufacturer (9) evaluates the patterns in the information sheet repository (8) to determine which endpoint which point or which related group of endpoints corresponds to a tool definition It defines each endpoint as an independent tool in an application model. It can be created, or in another implementation form, endpoints clustered around the same data entity. The points can be grouped under a single combined set of tools. In the information graph repository (8) function descriptor (14), endpoint name, path pattern, service name, extracted entity name and It is derived using the specified operation semantics. Arguments (8) in the information graph store schema (15), path parameters, query parameters, request body fields, mandatory and optional It is created by extracting field types from related fields. Information diagram 30 The return structure (16) in the repository (8) is used using the response body schema and entity fields. It is determined. The natural language description (17) in the knowledge graph repository (8), the purpose of the endpoint, The template is created by considering the associated data entity, transaction semantics, and graph context. It can be generated using a naturalistic, rule-based, or large language model-supported method. 7 Language description (17) is derived not only from endpoint names but also from the traffic collection module (5). LLM tool definition of contextual data and workflow patterns in the obtained audit traces It is derived by semantic analysis by the manufacturer (9). LLM vehicle definition The manufacturer (9) also trusts the types of parameters and field mappings to be included in the vehicle definition. being able to generate scores, limited publication of low confidence parameters, additional validation or 5 It can link user approval to certain conditions. This eliminates ambiguity within the system. opening up areas to direct and uncontrolled access to large language models is being blocked LLM work environment (10), where verified tool definitions are loaded and large language 10 The models process queries or tasks at runtime using these tools. It is the environment. New or updated tool definitions cause service interruption to the existing operating environment. It can be loaded without any additional effort. Thanks to this hot loading mechanism, both old and new vehicles can be loaded. Definition versions can be kept simultaneous for a certain period, backward compatibility during the transition process. It is protected and runtime calls can be routed to specific versions. 15 Autonomous knowledge graph and LLM vehicle definition generation method and system (1) continuous The inspection module (11) monitors traffic flow in real or near real time, and the information graph structural update required in the warehouse (8) or LLM vehicle designation manufacturer (9) It detects changes. The continuous monitoring module (11) detects previously observed endpoint 20 By storing structures, payload schemas, domain distributions, and reference summaries, a new endpoint the first time the endpoint is seen, a change in the parameter or response structure of an existing endpoint, Gradual deviations in the data payload structure, the emergence of a new entity relationship. or identifying situations where a previously unobserved data dispersion pattern is detected. Structural differences above the defined threshold value are shown in information graph 25. It automatically triggers the update and the associated tool definition production process. The verification and retrieval module (12) processes the generated new vehicle definitions in a multi-stage manner. This verifies the following: In the first stage, tool definitions undergo schema validity, argument type fit, and mandatory verification. in an isolated test environment in terms of the integrity of the areas and the consistency of the rotation structure is being verified. In the second stage, suitable vehicle definitions are sampled or controlled calls 30 It is tested through this. In the third stage, the use of vehicle definitions transferred to the live environment is examined. Success rates, error rates, and conformity to expected behavior are monitored. The error rate... exceeding a defined threshold, observing parameter mismatch, or unexpected operation. If time errors occur, the verification and rollback module (12) activates the system. 8 It automatically reverts to the previous stable version. Each vehicle definition version. They are numbered, stored immutably, and their version history is managed. In some implementations, the error rate, mismatch rate, or low confidence score is a specific When the threshold value is exceeded, the relevant tool definition, endpoint, or parameter set is temporarily disabled. It can be quarantined, released with limited access, or reverted to the previous stable version. 5 It can be rotated. One or more servers (2) to which one or more clients (3) are connected Autonomous information graph and LLM vehicle definition generation method and system used among (1), by one or more clients (3) or via inter-service communication target 10 Calls sent to systems (4) are collected by the traffic collection module (5). Traffic The parsing module (6) extracts the endpoints from these calls and their response logs. descriptors, payload schemas, response structures, correlation information, and persistent assets It extracts their identifiers. It analyzes existence and creation, reading, updating, and deleting. module (7) structures data entities on records, 15 within the system propagation and endpoint creation, reading, updating, deleting semantics It determines the nodes and edges in the information graph store (8). By representing it in this way, the current structural and semantic map of the system is created. Then, the LLM vehicle definition manufacturer (9) uses the structure in the information sheet repository (8), It creates tool definitions that can be used for large language models. Each tool definition has at least 20 a function identifier, argument scheme, return structure, and natural language description. It includes. The generated tool definitions are verified and undo module (12) by the schema. Testing for suitability, parameter compliance, sample call performance, and version reliability. The verified vehicle definitions are being processed. Service interruption to the LLM operating environment (10) It is loaded without any modifications and made available for use by large language model agents. 25 The continuous monitoring module (11) continues to track structural changes in the target systems (4). This involves new endpoints, changing data payloads, new data assets, and parameter differences. or it re-triggers the relevant modules when changes in data propagation are detected. Thus, the autonomous knowledge graph and the information in the LLM vehicle definition generation method and system (1) The diagram repository (8) and vehicle description are kept up-to-date. New 30 that are incorrect or incompatible. If definitions are detected, the verification and rollback module (12) reverts to the previous stable version. By doing so, it protects the safety and security of the working environment.

Claims

9 REQUESTS 1. Method and system for generating autonomous knowledge diagram and LLM vehicle definition (1), and its feature is; one or more servers (2), one or more servers connected to one or more servers (2) Multiple clients (3), target systems (4), microservice, application programming 5 raw request-response from target systems (4) such as interface gateway and databases traffic, event logs, and service interaction data in a unified observation layer. collecting traffic collection module (5), traffic collection module (5) obtained by traffic collection module (5) It processes raw footage, converts it into structured event logs, and delivers them to the endpoint. identifiers, HTTP methods or equivalent call types, path patterns, query 10 parameters, request payload schemas, response payload schemas, response structures, timestamps, correlation identifiers, trace identifiers, and persistent assets traffic parsing module (6) which extracts the identifiers, traffic parsing module (6) By using the persistent entity identifiers it generates, it identifies data entities and 15 instances of the same asset value at different endpoints, in different services, and in relation to time. By observing how they are seen throughout request-response sequences, it determines the spread of entities. Entity and creation, reading, updating and deleting analysis module (7), analyzed storing data in the form of nodes and edges, schema structures, and processing patterns. or context at the entity level that creates additional node types for event types. a queryable information graph store representing relationships and data flow (8), information 20 Using endpoint, entity, schema and relationship information stored in the diagram repository (8), Schemas of structured tool definitions usable by major language models LLM vehicle definition producer (9), in which case verified vehicle definitions are loaded and large language models that process their queries or tasks using these tools LLM work environment (10), by monitoring traffic flow in real or near real time, 25 information diagram repository (8) or LLM vehicle definition manufacturer (9) requires updating Continuous monitoring module (11) detects structural changes in the new vehicle produced Verification and rollback module that verifies definitions in a multi-stage manner (12), Network communication units (13) that provide data communication between modules, endpoint name, path 30 using the pattern, service name, extracted entity name, and defined transaction semantics. The function complement (14) in the derived information graph repository (8), path parameters, query parameters, request body fields, mandatory and optional (8) in the information graph repository formed by extracting the fields and field types. Using argument schema (15), response body schema and entity fields, the determined The purpose of the return structure (16) and endpoint in the information graph store (8) is to associate it with Template-based, rule-based, considering data assets, transaction semantics, and graph context. in the knowledge graph repository generated by a method based on or supported by a large language model (8) consists of natural language explanation (17).

2. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is; application programming interface gateway registers, reverse proxy logs, service network peripherals, application-level middleware logs, distributed Tracking data, message queue consumer listeners, event stream observation components. or performs data collection through similar sources and targets 10 systems. (4) centrally or logically store raw data representing actual working behavior It is a traffic collection module (5) that collects traffic in a unified observation layer.

3. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is that it generates persistent entity identifiers in response or request bodies 15 Unique identifiers such as id, uuid, key, and slug that are repeated and suitable for distinguishing an entity. Statistical repetition analysis of fields, field pattern extraction, rule-based methods or with learning methods, detecting and reproducing the format pattern of field values, again frequency, inter-source consistency, temporal stability, and level of discrimination By using and normalizing data from different sources, the system detects 20 It transforms heterogeneous records into a common event format and analyzes them together. It is the traffic separation module (6) that enables it to be done.

4. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is the structured 25 provided by the traffic separation module (6). semantic relationships between data assets, services, and endpoints through records relationships, entity expansions, and the semantics of creating, reading, updating, and deleting. Detecting, creating, reading, updating, and deleting semantics are only available at the endpoint. not through superficial matching based on name or HTTP method, but belonging to the same entity. 30 statistical methods, rule-based inference mechanisms, embedding vector with approaches or machine learning classifiers, response difference technique implementing, creating the first appearance of a new entity identifier, an existing one Reading the unchanged rotation of the entity, observing differences in field values. 11 The update ensures that the entity in question is no longer returned in subsequent related requests. or interpreting its inaccessibility as a deletion operation, the removed endpoint. assigning confidence scores to semantics, entity relations, and parameter types, low Entities and creation, reading, that direct safe results to user review, The update and delete analysis module (7) is. 5 5. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is; which endpoints produced a particular data entity, which services it was transferred between, at what steps it was changed, or which endpoint retrospectively with path-based resource queries showing how much resources were consumed by the points. following, providing the necessary contextual ground for vehicle definition production, vehicle definitions 10 while being generated, it measures the statistical frequency of use of call sequences and only specific ones. sequences of operations that exceed a certain repetition threshold and form a consistent pattern within the system When converting to a tool, the information graph store filters out random or erratic calls. (8) is.

6. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. 15 (1) and its feature is; by evaluating the patterns (8) in the information graph store, which endpoint which point or which related group of endpoints corresponds to a tool definition each endpoint, as a defined tool in a particular form of application, determines the definition of an independent tool. when creating, another form of implementation clusters around the same data entity. LLM tool definition manufacturer (9) 20 which brings together the points under a single combined tool set It is the fact that.

7. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is; new or updated tool definitions to the existing working environment While loading without service interruption, old and new vehicle definition versions are specific. Keeping the duration synchronized, maintaining backward compatibility during the transition process, and operating time 25 (10) is the LLM working environment which directs calls to specific versions.

8. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is; previously observed endpoint structures, payload schemes, area By storing distributions and reference summaries, a new endpoint is seen for the first time. A change in the parameter or response structure of an existing endpoint will cause a 30 change in the payload structure. gradual deviations occurring, a new relationship of existence emerging, or previously Identifying situations where an unobserved data dispersion pattern is determined It is a continuous monitoring module (11). 12 9. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is; - Verification and retrieval module (12), in the first stage, vehicle definitions scheme validity, argument type conformity, integrity of required fields, and return structure Verification of consistency in an isolated test environment, 5 - In the second stage, sample or controlled calls for approved vehicle identifications. testing through, and in the third stage, vehicle definitions taken into the live environment. usage success rates, error rates, and conformity to expected behavior monitoring It includes the steps of the process. 10 10. This method complies with claim 9 and its characteristic is; - The error rate exceeding a defined threshold indicates parameter mismatch. or verification in case unexpected runtime errors occur. and the rollback module (12) automatically restores the system to the previous stable version. to make it turn, 15 - It numbers each vehicle definition version, stores it in an immutable format, and managing version history It includes the steps of the process.

11. The method is in accordance with claim 9 and its characteristic is: - In some implementation forms of the verification and rollback module (12), the error rate is 20 When the discrepancy rate or low confidence score exceeds a certain threshold, the relevant Temporarily quarantine the vehicle definition, endpoint, or parameter set. to be able to obtain - Ability to release with limited access or revert to the previous stable version. rotatable 25 It includes the steps of the process.

12. Method and system for generating autonomous knowledge graph and LLM vehicle definition according to Claim 1. (1) and its feature is; - The traffic parsing module (6) which collects the calls sent to the target systems (4), From these calls and their corresponding response records, endpoint identifiers and data 30 load diagrams, response structures, correlation information, and persistent assets removing its identifiers, - The asset and creation, reading, updating and deleting analysis module (7), structured records of data entities, their presence within the system 13 propagation and endpoint creation, reading, updating, deleting semantics determination, - The results obtained are stored in the information graph repository as nodes and edges (8). by representing the system and creating its current structural and semantic map, - LLM vehicle definition is made by the manufacturer (9) using the structure in the information sheet repository (8), 5 It creates tool definitions that can be used for large language models, and each tool The definition must include at least one function identifier, argument scheme, return structure, and natural including a language explanation, - The generated vehicle definitions are schematized by the verification and rollback module (12). suitability, parameter compliance, sample call performance, and version reliability 10 testing in terms of, - Verified vehicle definitions, service interruption to LLM work environment (10) without loading and making it available for use by large language model agents. bringing, - The continuous monitoring module (11) tracks structural changes in the target systems (4) 15 This continues with new endpoints, evolving data payloads, and new data assets. When parameter differences or changes in data spread are detected, the relevant Re-triggering of modules, information graph store (8) and vehicle definition continuously keeping up to date, - In case of detection of erroneous or incompatible new definitions, verification and rollback will be carried out within 20 days. By reverting to the previous stable version of the module (12), the working environment preserving reliability It includes the steps involved in the process. 30

Images