Year: 2024
Type: article
Abstract: Extract, Transform, Load (ETL) pipelines are the lifeblood of enterprise data ecosystems, yet they remain highly vulnerable to silent failures, schema drift, and performance bottlenecks. Traditional monitoring approaches—based on static thresholds and reactive alerts—struggle to keep pace with the scale and complexity of modern data operations. This case study explores how predictive machine learning (ML) models can be embedded into ETL workflows to proactively identify, diagnose, and reduce failures before they impact downstream analytics.
Authors: Jonathan M., Lee, Rachel A., Martinez
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7081167523
