Year: 2024
Type: article
Abstract: This paper presents a comprehensive framework for modern API design that addresses the challenges of integrating machine learning operations into traditional and contemporary API architectures. The proposed approach combines three key elements: an AI-first architecture design for efficient vector operations and model serving capabilities alongside traditional data operations, event-driven patterns that enhance ML workflows and standard request-response interactions, and zero-trust security principles adaptable to both ML workloads and conventional API usage. The research demonstrates how these architectural patterns can be effectively implemented to create APIs that support traditional web services and modern computational workload operations while maintaining system scalability, security, and performance.
Author: Ganapathy Subramanian Ramachandran
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4406960556