Systems | Development | Analytics | API | Testing

Generative AI can produce tests instantly, but speed alone does not guarantee quality or safety. Without proper validation, AI-written tests can become brittle, redundant, or misleading, creating a false sense of coverage. This session looks at the risks of relying on AI-generated tests without the right controls in place.

Leadership today is facing a serious gap. Many people step into leadership and management roles without a clear understanding of how to lead, manage, or genuinely support their teams. While the fundamentals of leadership are simple, poor leadership makes it necessary to revisit the basics. Drawing from experience as a consultant, manager, individual contributor, and leader, this session focuses on what truly works and what consistently fails.

In machine learning, an 85% accurate model is often considered a success because we accept that data is messy, the real world is unpredictable, and chasing perfection is rarely worth the cost. However, in software testing, especially in the GenAI era, the question of “Can we test 100%?” still comes up. With AI systems producing probabilistic outputs where the same input can lead to different results, absolute coverage is unrealistic. Confidence scores already tell us that uncertainty is part of the system, and testing needs to acknowledge that reality.

Message brokers are a critical component of modern distributed systems, facilitating asynchronous communication between services. Load testing message broker integrations requires special considerations since the interaction patterns differ from traditional HTTP-based APIs. Speedscale provides specialized tooling to help you load test applications that integrate with message brokers by.