Year: 2025
Type: article
Abstract: Although machine learning (ML) has transformed many sectors, issues with scalability, robustness, and maintainability are frequently encountered during deployment and upkeep.To ensure that AI systems are durable, scalable, and maintainable, software engineering concepts must be incorporated into the creation of machine learning pipelines.In the context of developing machine learning pipelines, this study examines many software engineering techniques, including version control, modular design, testing methodologies, and continuous integration/continuous deployment (CI/CD). General TermsScalability, maintainability, and efficiency are improved in machine learning systems when software engineering ideas are incorporated into pipeline construction.For flexibility, use modular design; for traceability, use version control; and for automated testing and deployment, use continuous integration/deployment (CI/CD).
Authors: Sambedana Lenka, Suryasmita Sahoo, Rajesh Sahoo
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4409840687