AI Virtual Assistance Pipeline Studio for Automation Development

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Developing machines with Artificial Intelligence (AI) capabilities, such as machine learning (ML) and natural language processing (NLP), requires complex programming and a highly trained workforce, making the process difficult and time-consuming.

Innovation Solution

An AI-based virtual automated assistance system that provides an intelligent software framework for automating tasks by analyzing component processes, creating and executing pipelines, and integrating ML and NLP capabilities, allowing for minimal human intervention through a plug-and-play architecture and interactive GUI for building and executing services.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If complex programming and highly trained workforce are used to develop AI capabilities, then the quality and capability of AI systems are improved, but the development time and complexity increase significantly

Engineering Contradiction:
ImproveAI system capabilityVSAvoiddevelopment complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary platform that mediates between the complex AI development process and the end-user application. This platform provides pre-built AI services, templates, and tools that simplify the development process while maintaining access to sophisticated AI capabilities, thereby reducing development complexity without sacrificing system capability.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the AI development process into modular components and services that can be independently developed, configured, and deployed. This segmentation allows complex AI functionality to be broken down into manageable units, reducing overall development complexity while preserving the capability of the complete system.

Inventive Principle:
Principle #1Segmentation

2Reliability

If complex programming and highly trained workforce are used to develop AI capabilities, then the quality and capability of AI systems are improved, but the development time increases significantly

Engineering Contradiction:
ImproveAI system capabilityVSAvoiddevelopment time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by providing pre-configured AI services, templates, and frameworks that are prepared in advance. This allows developers to start with ready-made components rather than building everything from scratch, significantly reducing development time while maintaining access to sophisticated AI capabilities through the pre-prepared infrastructure.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If traditional AI development approaches are used, then sophisticated AI capabilities are achieved, but human intervention and expertise are required throughout the process

Engineering Contradiction:
ImproveAI capabilityVSAvoidautomation level
Core Design Contradiction:
ReliabilityVSExtent of automation

Solution Approach 1:

The patent enables self-service through automated workflows, self-configuring systems, and intelligent agents that can perform development tasks autonomously. The system provides self-service capabilities including automatic model selection, parameter optimization, and deployment, reducing the need for continuous human intervention while maintaining sophisticated AI capabilities.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS10725827B2Artificial intelligence based virtual automated assistance
Publication Date: 2020.07.28 ACCENTURE GLOBAL SOLUTIONS LTD
  • US10725827B2 patent drawing
  • US10725827B2 patent drawing
  • US10725827B2 patent drawing

AI summary

An Artificial Intelligence (AI) based virtual automated assistance system provides services pertaining to component processes of a task that is to be automatically executed. The virtual automated assistance system includes a pipeline studio that enables generating the services. Historical data pertaining to a service is accessed for training and validating various ML models. The ML models are scored and a selected ML model is registered as a service on the virtual automated assistance system. The services thus registered are represented as process blocks within the pipeline studio wherein the process blocks pertaining to the component processes of the task are arranged in order to form a pipeline. The pipeline thus constructed enables automatic execution of the task by receiving and processing a request pertaining to the task via the services that form the pipeline.