Sign In
Register

Partner with us

Register now

Call to action
Your text goes here. Insert your content, thoughts, or information in this space.
Button

Back to speakers

Mayank
Agarwal
Director of Product Management, AI Products
RingCentral
Mayank is Director of Product Management for AI Products at RingCentral, where he leads the AI Assistant products across the company's collaboration, communication, and contact center suites, shaping how millions of enterprise users interact with AI in their daily workflows. His work sits at the intersection of agentic AI architecture and large-scale product delivery, navigating the real-world constraints of latency, reliability, cost, and trust that separate impressive demos from systems businesses actually depend on. Before RingCentral, Mayank founded an AI-first global macro trading firm where he designed and built autonomous trading systems across equities, fixed income, commodities, and currencies. There, he built multi-agent architectures where specialized agents collaborated, competed, and coordinated with each other. He also pioneered reflection agents that continuously evaluated their own strategy performance, detected signal decay, and autonomously improved without human intervention. The unforgiving feedback loop of financial markets, where poor AI decisions have immediate, measurable consequences, shaped his approach to building systems that must be not just intelligent, but self-aware and resilient.
Button
15 April 2026 12:00 - 12:45
Panel: The blueprint to avoid obsolescence - building an architecture of defensibility
In the early days of the AI boom, "speed to market" was the only metric that mattered. This led to an explosion of "AI Wrappers"—products that essentially resold foundation model intelligence with a thin UI layer. But in 2026, the wrapper era has collapsed. As incumbents like Microsoft and Google bake those same features into the OS, PMs face an existential question: Is your product a feature, or is it a business? This panel explores the architecture of defensibility. We move beyond the "Build vs. Buy" debate to discuss the "Wisely Built" middle ground. Our experts will dive into how to move from "renting intelligence" to "owning the moat" through proprietary data flywheels, domain-specific fine-tuning, and vertical integration. If you are worried that your product is one API update away from obsolescence, this session will provide the blueprint for building a technical stack that actually belongs to you.