As AI moves from experimental to production systems, testing can no longer stop at accuracy and performance. Quality Engineers now play a pivotal role in ensuring that AI technologies are safe, transparent, and ethically aligned. This session explores how to extend QE practices into the AI lifecycle validating data, detecting bias, testing model behavior, and enforcing guardrails for fairness and reliability. You’ll learn practical strategies for embedding quality gates into ML pipelines, building explainability into automation, and monitoring AI drift in production. Join to discover how technical testers and engineers can become guardians of ethical AI using code, tests, and observability to earn user trust in intelligent systems.

• AI Testing Beyond Accuracy: Learn how to extend functional and non-functional testing to include ethical and behavioral dimensions of AI systems.
• Quality Gates for AI Pipelines: Discover how QE practices (data validation, model auditability, and bias checks) integrate into CI/CD for ML.
• Bias Detection and Explainability: Explore practical techniques and tools for surfacing hidden bias and improving model transparency.
• AI Safety in Production: Understand how to monitor AI drift, enforce human oversight, and test guardrails for real-world resilience.
• Ethics as an Engineering Practice: See how engineers can operationalize ethical principles using test automation, governance frameworks, and collaboration across teams.

In today’s software landscape, automated testing is no longer optional. The challenge is that traditional automation strategies, built around monoliths and simple pipelines, collapse when faced with dozens of microservices, asynchronous message buses, and environments that shift by the hour. In my talk, I will show how to address this problem by building scalable end-to-end automation frameworks composed of six key components: testware management, a domain-specific language, resource management, reporting, orchestration, and testability hooks in the system itself. Along the way, I will share how declarative DSLs reduce fragility, how dynamic resource allocation prevents “works on my machine” failures, and why observability must be treated as a design contract between developers and testers.

Attendees will leave with concrete takeaways:
how to think about automation as a product rather than a script, how to design frameworks that scale as fast as their systems do, how to apply these principles immediately to reduce flakiness, improve feedback loops, and ultimately deliver safer software faster. They will also learn how to position automation as a strategic enabler that bridges engineering practices with business outcomes, securing organizational buy-in for long-term quality investments.