Securing End-User Computing and Rogue Automation
Problem Statement
End-User Computing (EUC) and rogue automation processes are being used to
- Generate reports for regulatory compliance or management purposes.
- Support critical business functions and improve operational efficiency.
While these solutions offer quick wins, they pose significant risks, including:
- Security vulnerabilities due to lack of formal controls.
- Maintainability issues as they often lack standardized processes.
- Regulatory compliance risks as these processes are not managed by IT departments.
Scope
The EUC/Rogue Data Automations Analyzer Co-Pilot automates the assessment and migration of these ad-hoc solutions into more controlled IT environments, ensuring compliance and reducing risk.
Automated Assessment
Identify Data Sources
Automatically detect the data sources used within EUC applications.
Map Data Transformations
Trace and document data transformations performed by the EUC solutions.
Trace Data Lineage
Understand the flow and lineage of data throughout the EUC process.
Run Book Generation
Document EUC Functionality
Provide a clear summary of the purpose and key functions of each EUC.
Detail Data Flow
Map out the data flow, transformations, and dependencies.
Highlight External Factors
Identify dependencies or external factors influencing the EUC.
IT Product Backlog Creation
Based on the analysis, the system automatically generates a prioritized product backlog for IT teams, outlining the infrastructure and services needed to migrate the EUC. This includes:
Data Lake Requirements
Specifications for storing and managing EUC data in a structured environment.
Model Execution Platform
Infrastructure needed to run and manage EUC processes in an IT-managed environment.
Additional Infrastructure
Any extra services or platforms required to support the migrated EUCs.
Solution Overview
The EUC Analyzer Co-Pilot uses Azure OpenAI’s GPT-4, Python, Langchain, and AutoGen technologies to automate the process of assessing EUC solutions and rogue automation applications.
High-Level Process
Input
Accepts Tableau Prep Workflow (TFL) files and other EUC formats as input.
Analysis
- Uses Prompt Engineering to query Azure OpenAI GPT-4 for insights into data sources, transformations, and data flow.
- Extracts key technical details from the workflows to provide comprehensive analysis.
Integration
- Langchain: Chains multiple LLMs together for a deeper technical understanding.
- AutoGen: Utilizes text generation to create detailed assessment reports and run-books.
Output
- Generates comprehensive assessment reports that include:
- Data sources used within the EUC workflow.
- Detailed data transformations and manipulations.
- Potential areas for improvement, optimization, or risk mitigation.
Benefits
Streamlined Migration
Automates the assessment and documentation process, significantly reducing the time and effort required to migrate EUCs to an IT-managed environment.
Improved Visibility
Provides clear visibility into data sources, transformations, and dependencies, helping organizations mitigate risks posed by rogue automation processes.
Informed IT Backlog
The automatically generated backlog helps IT teams prioritize tasks and allocate resources effectively for EUC migration and compliance initiatives.