Modern software systems have become increasingly complex, making software testing and quality assurance more challenging, time-consuming, and costly when relying on traditional manual approaches. Recent advancements in Large Language Models (LLMs) have introduced new opportunities for intelligent software testing through their capabilities in natural language understanding, code generation, reasoning, and knowledge extraction. This study proposes an Intelligent Test Engineering Framework that leverages LLMs to automate test generation and defect triage across the software development lifecycle. The framework integrates requirement analysis, automated test case generation, test script creation, defect classification, severity prediction, and rootcause recommendation into a unified quality engineering pipeline. It consists of five key components: the Requirement Intelligence Engine (RIE), Automated Test Generation
Module (ATGM), Test Execution Support Layer (TESL), Intelligent Defect Triage Engine (IDTE), and Continuous Learning Repository, which collectively process software artifacts, generate testing assets, analyze defects, and provide actionable recommendations for Software Development Engineers in Test (SDETs). Experimental evaluation using software project datasets and simulated testing environments assessed metrics such as test generation accuracy, defect classification accuracy, triage effectiveness, and automation success rate. Results demonstrated significant improvements in testing productivity, defect resolution speed, test coverage, and overall software quality assurance. Automated test generation enhanced coverage, while AIassisted defect triage reduced analysis time and improved prioritization accuracy,
highlighting the potential of LLMs to transform conventional testing workflows into intelligent, adaptive, and highly efficient quality engineering systems.
Keywords : Large Language Models, Software Testing, Test Automation, Defect Triage, Quality Engineering, Artificial Intelligence, SDET, Generative AI.
Author : Rajasekhar Sunkara Senior Software Development Engineer in Test (SDET) | Quality Engineering Specialist,Austin, Texas, USA
Title : Intelligent Test Engineering with Large Language Models: A Framework for Automated Test Generation and Defect Triage
Volume/Issue : 2023;05(01)
Page No : 53-73