LLM-Assisted Task Mining for Automated Workflow Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional task mining is time-consuming and requires significant manual intervention, such as manually recording and merging workflows, which is inefficient.
Innovation Solution
Utilizes large language models to automatically generate and modify structured definitions of workflows based on natural language descriptions and API descriptions, reducing the need for manual steps and enhancing workflow capture and retrieval efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional task mining is performed by manually recording user interactions, then workflow data can be captured, but the process is time-consuming and requires significant manual intervention
Solution Approach 1:
The system performs automatic task mining by capturing user interactions through recording mechanisms without requiring manual intervention. The workflow data is collected and processed automatically, allowing the system to serve itself in capturing and analyzing task information.
Solution Approach 2:
The patent replaces manual mechanical recording processes with automated computational systems. Large language models and machine learning algorithms substitute for human operators in capturing, processing, and analyzing workflow data, eliminating the need for manual task mining operations.
2Productivity
If conventional task mining requires manual merging of workflow records, then workflow integration can be achieved, but the process is inefficient and time-consuming
Solution Approach 1:
The system automatically merges workflow records by capturing multiple variations of workflows and integrating them through automated processes. The large language models perform the merging operation without human intervention, allowing the system to handle workflow integration independently.
Solution Approach 2:
The patent combines multiple workflow records and variations into a unified workflow structure. The system merges captured interactions from different sources and time periods, integrating them into coherent workflow models through automated analysis and synthesis.
3Loss of time
If structured workflow definitions are generated automatically using large language models, then workflow capture time is reduced, but the system complexity increases
Solution Approach 1:
Large language models serve as intermediaries between raw user interaction data and structured workflow definitions. The LLMs translate unstructured captured interactions into organized workflow representations, bridging the gap between data collection and structured analysis without requiring complex manual processing.
Solution Approach 2:
The system transforms workflow data from unstructured captured interactions into structured definitions by changing the organizational parameters. Large language models restructure the data format, converting raw interaction sequences into standardized workflow models with defined nodes, edges, and metadata.
Data Source
AI summary
Systems and methods for generating a structured definition of a graph of a workflow are provided. 1) a natural language description of a workflow and 2) a description of one or more APIs (application programming interfaces) for performing the workflow are received. A structured definition of a graph of the workflow is generated using a language model based on the natural language description of the workflow and the description of the one or more APIs. One or more nodes in the structured definition of the graph of the workflow is mapped to at least one of the one or more APIs. The structured definition of the graph of the workflow is output.