Data-Driven API Test Case Generation Using AI and Model Context Protocol

 In the rapidly evolving landscape of software development, Application Programming Interfaces (APIs) are critical for enabling seamless integration and functionality across diverse systems. Ensuring their reliability and robustness is essential, yet traditional testing methods often struggle with efficiency and comprehensive coverage, particularly in addressing complex interactions and edge cases. This paper presents an innovative approach that leverages Artificial Intelligence (AI) and Model Context Protocol (MCP) to automate the generation of test cases for APIs. By employing advanced machine learning algorithms, and MCP interfaces, our system analyzes API specifications, usage patterns, and historical test data to intelligently generate a diverse and thorough set of test cases.

This AI-driven methodology accelerates the testing process and enhances coverage by identifying and addressing edge cases that manual testing might overlook.

Paper | Presentation

Harold Wilson

Harold is a distinguished software quality assurance professional with extensive experience in building and leading QA teams across diverse industries, including aerospace, digital media, and semiconductors. His career highlights include serving as a Test and Reliability Consultant for the United States Space Force, where he ensured the dependability of mission-critical software systems, and as Director of Quality Assurance at Entercom Digital (radio.com), where he oversaw the quality of widely used digital platforms. With expertise in software testing, reliability engineering, embedded systems, and IT security-including compliance with PCI and HIPAA standards-Harold has consistently developed robust QA processes from the ground up. He began his career as an Electronics Technician in the United States Navy, a Gulf War veteran, and holds a Bachelor's of Science in Computer Science from the College of Santa Fe, Albuquerque Campus, where he also served as adjunct faculty, teaching computer science courses. Harold's deep understanding of software quality challenges across multiple domains positions him as a valuable contributor to discussions on innovative testing strategies and reliability practices. Bachelors of Science in Computer Science from the College of Santa Fe, Albuquerque Campus where he also served as adjunct Faculty.

Joseph Petsche

Joseph Petsche is a seasoned Software Architect with over two decades of experience, specializing in automation, integration, and quality assurance. Currently serving as an Automation Architect at EverDriven, he streamlines DevOps pipelines and fosters cross-team collaboration to deliver high-quality software solutions. His notable career includes automating data flows at WebMD Health Services, building resilient software through automated testing at Red Rock Tech Solutions, and contributing to community engagement as a Board Member at Evergreen Curling. With expertise in C#, Powershell, and Playwright, Joseph is passionate about solving complex problems and ensuring client satisfaction. An avid mountain climber, he approaches challenges with strategic determination. Joseph holds a degree in Computer Science & Engineering from the University of Toledo.