AI Workflow Generation via Data Map Path Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional business workflow management systems are inefficient in processing data changes and responding to them in a timely manner, relying heavily on user experience and being inflexible across different enterprises and business projects, with existing systems struggling to adapt to various business scenarios effectively.
Innovation Solution
A data-based workflow generation device and method that uses a processor and storage device to search and execute a data map to find and execute the best task path, repeatedly adjusting the input data set until it matches a target data set, enabling automatic generation and execution of task data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional business workflow management systems are used, then business processing can be performed through forms and workflows, but the systems cannot effectively perceive data changes and respond in a timely manner
Solution Approach 1:
The patent replaces conventional mechanical workflow systems with an AI-based system that uses large language models to automatically generate workflows. The system substitutes traditional form-based manual processing with intelligent agents that can autonomously perceive data changes, analyze requirements, and execute tasks, thereby dramatically improving response speed and adaptability to data changes.
Solution Approach 2:
The system enables self-service by allowing the AI agent to autonomously generate workflows, select appropriate tasks, and execute operations without human intervention. The agent can independently perceive data changes, query information from information carriers, and complete business processes, eliminating the need for manual workflow configuration and execution.
2Extent of automation
If conventional business logic is used to assist manual data processing, then user operations are supported, but a large number of operations still need to be initiated by users
Solution Approach 1:
The system implements self-service by enabling the AI agent to autonomously initiate and execute operations based on perceived data changes. The agent can independently determine when workflows need to be triggered, select appropriate tasks from the data map, and execute them without requiring user initiation, thereby significantly increasing automation while maintaining ease of operation through natural language interaction.
Solution Approach 2:
The system incorporates feedback mechanisms where the AI agent continuously monitors data changes, analyzes the impact of these changes, and automatically adjusts workflow execution accordingly. The agent receives feedback from data sources, information carriers, and execution results, enabling it to autonomously initiate operations and adapt to changing business requirements without user intervention.
3Adaptability or versatility
If conventional systems establish complete sets of systems according to different enterprise systems and business projects, then specific business needs are met, but the systems cannot be flexibly applied to different enterprises or business projects
Solution Approach 1:
The patent implements universality by creating a unified AI-based workflow generation system that can serve multiple enterprises and business projects through a single platform. The system uses a universal data map structure and large language model that can adapt to different business scenarios without requiring separate system configurations, enabling one system to fulfill multiple functions across diverse organizational contexts.
Solution Approach 2:
The system achieves adaptability through parameter changes by allowing the AI agent to dynamically adjust workflow parameters, task selections, and execution strategies based on the specific enterprise context and business project requirements. The underlying system architecture remains constant, but the operational parameters are flexibly modified through natural language processing and AI reasoning to suit different scenarios.
4Loss of information
If conventional process engines or ERP systems are used, then business workflow management is provided, but the systems have insufficient carrying capacity for knowledge and rely on operator experience
Solution Approach 1:
The patent replaces conventional process engines and ERP systems with an AI-based workflow generation system that uses large language models. This substitution dramatically increases knowledge carrying capacity by enabling the system to store, retrieve, and apply business knowledge, best practices, and domain expertise encoded in the data map and processed by the AI agent, eliminating reliance on individual operator experience.
Solution Approach 2:
The system introduces an AI agent as an intermediary between data sources and workflow execution. This intermediary possesses high knowledge carrying capacity, able to query information from multiple information carriers, synthesize knowledge from diverse sources, and make informed decisions about workflow generation and task execution, thereby improving judgment accuracy beyond human operator capabilities.
Data Source
AI summary
A data-based workflow generation device and method thereof are provided. The data-based workflow generation device includes a storage device and a processor. The processor is coupled to the storage device. The processor receives a data set. The processor takes the data set as an input data set, and searches a data map according to the input data set to obtain a best task path. The processor executes the best task path to obtain an output data set. The processor takes the output data set as the input data set, and repeatedly searches the data map according to the input data set, and executes a corresponding best task path until the output data set is the same as a target data set.