Strategies for Testing AI-Based Systems

Description

The rapid proliferation of Artificial Intelligence (AI) and autonomous agentic systems presents unique and complex challenges for traditional software testing practices. Testers must evolve their skills to evaluate the quality, reliability, security, and ethical behavior of these intelligent systems.

This hands-on course equips testers with essential techniques and strategies for effectively testing AI and agentic AI systems. Participants learn how to apply specialized testing methods, tools, and workflows to validate model and solution behavior, from planning and execution through automation and reporting.

Key takeaways from this class include:

• Understanding the key differences between traditional software testing and AI systems testing
• Defining quality metrics and test strategies for ML models and AI data pipelines
• Testing agentic AI behavior, including goals, plans, tool use, and emergent outcomes
• Performing adversarial and robustness testing for AI components
• Analyzing and reporting ethical and safety aspects of AI performance
• Leveraging tools for data quality analysis, explainability (XAI), and continuous validation

Who Should Attend

This course is ideal for software testers, quality assurance engineers, and test managers who validate systems that include AI, machine learning, or agentic components. A foundational understanding of software testing principles and high-level AI/ML concepts is recommended.

Laptop and RDP Required

This class includes hands-on activities using sample software. Each attendee should bring a laptop with a remote desktop protocol (RDP) client pre-installed. Connection details and credentials are provided during class. Please coordinate with your IT administrator beforehand to confirm your RDP client can access a virtual machine in an AWS environment.