AI Agents in Action: Autonomous Test Generation, Automation & Defect Management
Software testing is no longer just about generating test cases - it's about intelligent automation that learns, adapts, and evolves. While AI-driven test generation using tools like ChatGPT or AWS Bedrock is becoming standard, these models often function in isolation, producing test cases based only on the immediate input they receive. But what if AI could go beyond one-time generation and instead act as an autonomous agent, continuously refining, executing, and even managing defects throughout the testing lifecycle
This workshop introduces the power of Agentic AI - AI-driven agents that don't just generate test cases but perform goal-oriented tasks such as:
- Writing test cases based on user stories
- Reviewing and optimizing test coverage across past and present releases
- Automating script creation tailored to specific frameworks
- Identifying defects and even creating defect reports based on execution insights
By integrating GenAI APIs with Vector Databases, these agents develop a memory, enabling them to retrieve past test cases, learn from previous defects, and dynamically adapt test automation strategies.
Through hands-on demonstrations, we'll showcase how to design and implement intelligent agents that can streamline your entire testing pipeline - from test case generation to execution, defect detection, and reporting.
Jigesh Shah
Jigesh Shah is an Engineering Manager(QA) with rich experience in projects for clients across eLearning and HealthCare industries. He has been extensively involved in test solutions development where he has developed many reusable testing tools, frameworks and templates. He provides testing solutions to clients to improve the quality of testing in his current role at Deloitte. He played a key role in Test advisory, Test management, test planning and execution for the US and global clients and led the offshore testing effort. He has extensive experience Performance Engineering, SQL optimization and tuning and system resource monitoring using multiple tools.