Speak at The Fifth Elephant 2026 Annual Conference

The “why”?

Every decade or so, a protocol comes in and changes everything. The Model Context Protocol (MCP) was one such thing which took the place of this decade and which became the momentum to AI. Since its first specification release, it has fundamentally reshaped how AI models talk to the world: external tools, live data, real services. But here’s what nobody tells you about the quiet revolution it has gone through.Ever wondered why MCP was born as a stateful, connection-oriented protocol, and what forced it to reinvent itself into the lean, stateless architecture you see in the latest release candidate? Ever wondered what it actually takes for an entire industry to rally behind a protocol; the politics, the pain points, the tipping moments? We’ll answer both.

This session is a front-row seat to that evolution. We walk through every major version of the MCP spec, the design decisions that seemed brilliant at the time, the community pushback that forced a rethink, and the real-world implementation battles that separated good ideas from great ones. We’ll unpack what the shift to stateless architecture means in practice: how it changes the way you design servers, how it unlocks scale that wasn’t possible before, and how the developer experience transforms when the protocol gets out of your way. You’ll leave with a mental model of where MCP came from, why it changed, and exactly what it means for building AI-integrated systems that are built to last.

Takeaway

A clear understanding of how and why the MCP spec even started to begin including the specific pain points that motivated the transition from stateful to stateless architecture so you can make informed design decisions in your own MCP-based integrations.
Practical insight into migrating or building MCP servers under the latest spec, including what breaks between versions and how to handle backward compatibility gracefully.

Audience

This session will be most beneficial for AI/ML practitioners building or integrating LLM-powered applications, platform and infrastructure engineers designing tool-calling or agent orchestration systems, and developer tools enthusiasts who want to stay ahead of the evolving AI tooling ecosystem.

Bio

Hritik Raj is a ML Systems Engineer at Nutanix, where over the past two years he has led the team’s work on LLM observability and production MCP adoption. He’s an active contributor to the Kubernetes-native Envoy AI Gateway project, focused on making agentic AI secure and observable at enterprise scale.

Drive link of the slides - here