Most enterprise AI agents fail quietly, completing tasks with no awareness of their own errors. At my quantitative trading firm, I built self-improving AI systems where failure had immediate financial consequences, forcing architectures that detected mistakes, diagnosed root causes, and improved autonomously. At RingCentral, I've applied these same patterns to our AI Assistant, now serving millions of enterprise users. The core insight: self-correcting agents aren't a research problem, they're a product design problem. We'll explore the framework refined across both domains: how to build AI agents that know when they're wrong, the architecture patterns that make self-correction reliable in enterprise environments, and how to measure autonomous improvement over time. Attendees will leave with a reference architecture and a KPI framework for evaluating self-improving agents, ready to apply to their roadmaps next quarter.

This is your chance to meet the room, share your stories through our ice breakers and prepare for the day ahead.

Day two kicks off with our chairperson's opening remarks, It's time to get moving and networking! Let's dive into the topics of the day and meet some new faces.

Our chairperson will close out the end of day one of the summit as we reflect on all the sessions of the day and prepare for day two.

Every product leader knows that unchecked legacy products and mounting technical debt act as a "hidden tax," draining engineering capacity and crippling innovation velocity. This panel tackles the political, organizational, and financial gravity of portfolio rationalization. We will discuss and debate the strategic frameworks required to set clear, data-driven sunset criteria for profitable but stagnant products. Attendees will gain actionable insights on how to quantify technical debt as a business risk, secure alignment with finance and engineering leaders, and effectively reallocate freed-up resources to fund the next generation of growth products and competitive AI initiatives.

Our chairman will close out the end of the summit.