Diagram-to-Workflow Conversion Using LLM Skeleton Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generating and modifying workflows in cloud-based architectures is a tedious and time-consuming process that leads to inefficient utilization of computing resources and errors, preventing workflows from executing to completion.
Innovation Solution
A workflow editing tool that utilizes large language models (LLMs) to convert diagrams into skeleton workflows, allowing for the definition of placeholder activities based on the diagram data, and subsequently generating complete workflows by specifying placeholder values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If workflow editing tools are used to generate workflows activity-by-activity, then workflows can be created with detailed control, but the process becomes tedious and time-consuming
Solution Approach 1:
The system performs preliminary actions by automatically generating a skeleton workflow structure and placeholder activities based on the uploaded diagram, eliminating the need for users to manually create each activity from scratch. The LLM analyzes the diagram and pre-populates the workflow with appropriate activities and their basic properties.
Solution Approach 2:
The system copies the structural information from the uploaded diagram and transforms it into a workflow format. The diagram's visual representation of processes and activities is replicated as a skeleton workflow structure, preserving the logical flow and relationships while adapting it to the workflow execution environment.
2Reliability
If manual workflow creation is performed, then customization is possible, but errors occur that prevent workflow execution
Solution Approach 1:
The system performs self-service by automatically analyzing the uploaded diagram and generating a complete skeleton workflow structure without requiring manual configuration. The LLM independently interprets the diagram and creates appropriate activities, connections, and properties, reducing human intervention and associated errors.
Solution Approach 2:
The system replaces the mechanical manual process of workflow creation with an automated intelligent system. Instead of manually configuring each workflow element, the LLM-based system automatically generates the skeleton workflow, substituting human manual operations with automated AI-driven processes that reduce errors.
3Productivity
If traditional workflow tools are used, then computing resources are available, but resource utilization is inefficient
Solution Approach 1:
The system applies partial action by generating only the essential skeleton workflow structure and placeholder activities needed to represent the diagram's core logic. Rather than creating a fully detailed workflow immediately, it produces a streamlined structure that can be further refined, reducing unnecessary computational overhead.
Data Source
AI summary
A method includes receiving data representative of a diagram, converting the data to a large language model (LLM) prompt, providing the LLM prompt to an LLM, receiving, from the LLM, based on the LLM prompt, a skeleton workflow, wherein the skeleton workflow includes a placeholder activity, and wherein the placeholder activity includes a placeholder value for a property of the placeholder activity, and generating a workflow based on the skeleton workflow and the data, wherein generating the workflow includes defining the placeholder value for the property of the placeholder activity based on the data.