Automation Engine Orchestration Using Generative AI and Existing RPAs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automations and artificial intelligence models are not effectively integrated to accomplish specific user goals, leading to unused automations and the need for new automations to be created for each task, thereby underutilizing existing automation capabilities.
Innovation Solution
An automation engine that integrates robotic process automations (RPAs) with generative artificial intelligence models to process natural language problem statements, determining an automation plan and executing existing automations to achieve specific tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional automations are used independently, then each automation can perform its specific function, but they cannot be effectively integrated to accomplish complex user goals, leading to underutilization of existing automation capabilities
Solution Approach 1:
The patent introduces an automation engine as an intermediary component that sits between the user and existing automations. This engine receives natural language problem statements, translates them into executable automation plans, and coordinates multiple existing automations to work together. The automation engine acts as a mediator that integrates disparate automation components without requiring complex direct connections between them, thereby improving adaptability while managing system complexity.
2Productivity
If new automations are created for each specific task, then the task can be accomplished, but existing conventional automations remain unused, wasting automation resources
Solution Approach 1:
The patent implements a universal automation engine that can handle multiple different tasks by coordinating existing automations in various combinations. Rather than creating specialized automations for each task, this single engine serves multiple functions by dynamically selecting and orchestrating appropriate existing automations based on the specific problem statement, thereby improving productivity while reducing automation resource waste.
Solution Approach 2:
The system performs preliminary action by pre-building a library of conventional automations that can be reused. The automation engine analyzes the problem statement and selects from this pre-existing library rather than creating new automations from scratch for each task. This approach ensures that existing automation resources are fully utilized before investing resources in creating new automations.
3Adaptability or versatility
If AI models create new automations for each task, then specific user goals can be achieved, but the numerous conventional automations remain unused
Solution Approach 1:
The automation engine performs self-service by automatically analyzing problem statements and determining which existing automations to deploy. Rather than requiring manual creation or selection of automations for each task, the system autonomously selects and coordinates the appropriate subset of existing automations from its library, reducing the quantity of automations needed while maintaining high adaptability to different user goals.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method is provided. The method is executed by an automation engine implemented as a computer program within a computing environment. The automation engine executing processing strategy based on a problem statement. The method includes receiving the problem statement comprising a specific requested task in natural language and automatically processing the problem statement by utilizing a generative artificial intelligence (AI) model to manipulate the natural language of the problem statement to create a new computer digestible script defining the specific requested task. The method includes automatically determining the processing strategy comprising an automation plan of computer generated steps that achieves the specific requested task and executing the plurality of existing automations according to the processing strategy.