Serverless Computing for Scalable Cloud Applications
Year: 2025
Type: article
Abstract: Serverless computing has emerged as a leading execution model for building scalable cloud applications, offering automatic provisioning, event-driven execution, and fine-grained billing. Unlike traditional cloud architectures that require continuous resource management, serverless platforms dynamically scale functions based on demand, allowing developers to focus on application logic rather than infrastructure. This paper examines how serverless architectures enable scalability, improve operational efficiency, and reduce cost for a wide range of application domains, including data processing pipelines, web services, and IoT analytics. The study evaluates key characteristics of serverless systems such as stateless function execution, cold-start behavior, and integration with managed cloud services. It also explores advanced techniques for optimizing function performance, including predictive autoscaling, workload-aware scheduling, and state management via service meshes and distributed storage. A comparative analysis across major platforms—AWS Lambda, Google Cloud Functions, and Azure Functions—highlights differences in execution environments, concurrency models, and scalability limits. The paper identifies challenges that inhibit broader adoption, such as vendor lock-in, debugging complexity, and limited support for long-running or stateful workloads. Experimental evaluations conducted on prototype applications show that serverless deployments can achieve near-linear scaling under bursty workloads while maintaining low operational overhead. The findings suggest that serverless computing is well suited for modern cloud-native applications but requires careful design to ensure cost predictability and performance consistency. The paper concludes with architectural recommendations and future research directions for building scalable and resilient serverless systems.
Author: Lara Mendez Ortiz
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7115690552