AI coding tools are now used by 90% of engineering teams, accelerating development velocity exponentially. However, this speed has introduced a not-so-hidden problem: we see an 8x increase in duplicated code, 100% increase in code churn and a clear decline in application quality. How do we build faster without creating fragile, untrustworthy products?

This session dives into the "AI Productivity Paradox" and presents a part of the solution in the form of Agentic Testing. We'll break down how Squidler’s AI agent autonomously tests applications from the user's perspective, moving beyond traditional code-based QA towards automated manual testing. You will learn how combining agentic, intent-based navigation with deterministic validation can create robust, fast, and simple quality assurance tailored for the future of software development.