Interpreting Graphs Generated from Metric, Log, and Stream Data Using Artificial Intelligence, and Implementing AI-Powered Business and Ticket Management Methods.

Abstract

The invention relates to a method for analyzing graphs generated from metric, log, and flow data collected from monitored systems using an artificial intelligence engine to identify anomalies, root causes, and performance bottlenecks; generating automated or manual work tickets and notifications based on the identified results; and identifying process and load bottlenecks through artificial intelligence in the work management layer based on the generated work tickets, and presenting action recommendations. Within this method, graphical data, represented as time series or data streams, is interpreted individually or in groups by an artificial intelligence engine supported by causal graphs and reciprocal information extraction (RAG). The root cause analysis output generated at the graphical level is converted into business records. At the business management layer, these business records are analyzed using language models and artificial intelligence techniques to identify load imbalances and process disruptions, and to suggest corrective actions.

Description

1 TARIFF Interpreting Graphs Generated from Metric, Log, and Flow Data Using Artificial Intelligence. AI-Powered Business and Ticket Management Method 1. Technical Area The invention derives from metric, log, and stream data collected from complex computing and operational infrastructures. Interpreting the generated graphs using artificial intelligence techniques, identifying anomalies and root causes; this Artificial intelligence in the business management layer through work records (tickets) created based on findings. This relates to the methodology of identifying process and load bottlenecks and proposing action plans. 2. Evaluation of Known Techniques In traditional monitoring and operational management systems, log, metric, and stream data are typically They are visualized as graphs on separate panels and manually operated by the operators. They are correlated. Analyzing graphs from different data sources with human intervention, 15 This can prolong decision-making times, lead to errors in anomaly and root cause detection, and trigger alarms. It increases your fatigue. Additionally, job tickets are created as a result of operational disruptions in existing structures. It is mostly used solely for status tracking purposes; the distribution of work records within the process, Resolution times and workload imbalances on personnel / units are analyzed using artificial intelligence. 20 This is not possible. Graphic-based root cause identification and process bottleneck analysis at the business management level. a structure known for presenting it as an integrated method and independent of infrastructure components It is not found in the technology. 3. Purpose of the Invention The invention aims to generate graphs from metric, log, and stream data using an artificial intelligence engine. reducing the need for human intervention by enabling automatic interpretation, anomaly and The goal is to speed up the identification of the root cause. Another purpose of the invention is to create automated or manually generated processes as a result of root cause identification. by applying artificial intelligence and language modeling techniques to the records (tickets) the load in the process Identifying imbalances and process bottlenecks and taking action to improve operational efficiency 30 The goal is to generate suggestions. Thus, data to graph can be created without being dependent on a specific hardware or field component. From graphical analysis to root cause, from root cause to job record, and from job record to AI-powered job A seamless chain of methods is provided, extending from management to bottleneck identification. 4. EXPLANATION OF THE FIGURES 35 Figure 1 – General Flowchart of the Method: Generating graphs from metric, log, and flow data. Interpreting graphical data with an artificial intelligence engine to identify root causes, and then implementing solutions based on the findings. Generating tickets and notifications, and identifying bottlenecks through artificial intelligence based on the generated work records. It shows the general flow, including the steps from identifying the problem to proposing action. 2 Figure 2 – Flowchart for Interpreting Graphical Data with Artificial Intelligence: Single or multiple. Providing graphical data as input to the artificial intelligence engine, causality graph, and Generating the root cause hypothesis and identifying the anomaly / root cause / action using alternating information inference (RAG). This shows the steps for submitting the proposal output to the business registration and notification system. Figure 3 – Two-Tier Business Management and Bottleneck Analysis Architecture: Graph 5 for Layer 1 Job records are created from data using AI root cause analysis, and job records are opened at Layer 2. process / load bottleneck detection via artificial intelligence and mutual reinforcement between the two layers It shows the flow. 5. Detailed Description of the Invention 10 The method described in the invention consists of two fundamental layers of artificial intelligence analysis: 1. Layer for Interpreting Graphic Data with Artificial Intelligence: Metric, log, and stream data collected from monitored systems will show temporal changes and relationships. These are generated as single or multiple graphs. The artificial intelligence engine processes these graphs visually and through data. It analyzes the structure in terms of its dimensions. Using the causal graph method, it examines the 15 differences between the graphs. Interactions and correlations are evaluated. Furthermore, the past is analyzed using a reciprocal information extraction (RAG) framework. Anomalies, root causes, and performance bottlenecks are identified by utilizing case data. The analysis results are converted into a structured report, notification, or business record (ticket). 2. AI-Powered Business Management and Bottleneck Detection Layer: Job records (tickets) created as a result of graphical analysis or manually entered into the system, job management 20 It is transferred to the next layer. The artificial intelligence engine and language model working at this layer analyze the types of work records, resolution times, assigned unit / personnel workload distribution, and waiting times in process steps It examines and analyzes load imbalances and process bottlenecks in operational processes. This is done. Process improvements or resource redistribution are implemented to address the identified bottlenecks. Action proposals are generated and presented to the relevant parties. 25 These two layers work in a mutually reinforcing manner: detected at the graphical level. While technical root causes trigger transaction logs, AI analysis performed at the transaction log level... It enables integrated operational improvement by identifying organizational or process bottlenecks. 6. Industrial Applicability The invention relates to information technology operations, data centers, telecommunication systems, 30 It can be directly applied in industrial monitoring infrastructures and enterprise process management software. It is capable of data monitoring and business management without any hardware or field node limitations. It can be integrated as a software and logical method into all systems that offer software. 35

Claims

3 REQUESTS 1. Graphical-based analysis of metric, log, and stream data in monitored systems and AI-powered workflows. It is a management method, its characteristic is; • At least one unique or multiple metric, log, and stream data obtained from the monitored systems. creating graphs, • The generated graphs are processed by an artificial intelligence engine to identify anomalies, root causes, and Identifying performance bottlenecks, • Job registration based on root cause and anomaly data obtained from the analysis. (ticket) and / or notification creation, 10 • In the business management layer, which is run through the generated business records, artificial intelligence is used. Identifying workload distribution and process bottlenecks, and proposing actions for the process. presentation A method characterized by its inclusion of several steps.

2. The method according to claim 1, its characteristic is; the artificial intelligence engine makes a distinction between single or multiple graphical data points. by establishing causal graphs and using vector database-supported information extraction (RAG). This involves generating a root cause hypothesis based on past cases.

3. A method according to claim 1 or 2, characterized by its graphical representation of metric, log, and stream data. Time series, system performance indicators, and data flow patterns are analyzed by an artificial intelligence engine. It is an integrated assessment. 20 4. This method, according to Claim 1, is characterized by its artificial intelligence analysis at the business management layer, for opened jobs. the assignment status of records, resolution times, and the distribution of workload on the workforce It involves identifying process bottlenecks by evaluating them.

5. The method, according to Claim 1, is characterized by its use of artificial intelligence root cause analysis at the graphical level. The bottleneck and recommendation analysis at the management level is a two-layered structure that feeds into each other. 25 It is the execution.

Images