Will MCP be the Future of Agents?
This meetup covers deep technical demos, lightning talks, and conversations about MCP (Model Context Protocol) and AI Agents!
You will hear the industrial insights from ReAPI, KONG, CNCF WasmEdge project, Gaia, LangDB, FatfishTechnology and others.
This FREE event is held at AWS’s office (Thank you, AWS and Gaia). Pizza and drinks will be provided courtesy of our sponsor ReAPI, Kong and Gaia.
Agenda
6:00 PM: Door opens
6:20-6:40 PM: Enjoy the Pizza
6:40-7:00 PM: Empowering Context-Aware APIs with Kong's MCP Server — Robin Cher, Kong
7:00-7:20 PM: Context-First APIs: Supercharging AI IDEs with the OpenAPI MCP Server — PeiSong Xiong, Founder of ReAPI
7:20-7:40 PM: My MCP server is an AI agent --- Michael Yuan, CNCF WasmEdge
7:40-8:00 PM: Beyond the Spec: Practical MCPs Through LangDB AI Gateway — Vivek Gudapuri, Co-Founder at LangDB
8:00-8:30 PM: Networking
The agenda is subject to change.
Details:
Context-First APIs: Supercharging AI IDEs with the OpenAPI MCP Server
AI tools are getting smarter—but most OpenAPI specs are still written for humans, not machines. In this talk, we’ll explore a “context-first” approach to API design that goes beyond traditional docs. You’ll see how our OpenAPI MCP server turns static specs into searchable, AI-ready context that powers smarter autocomplete, Q&A, and code generation inside your IDE.
We’ll cover: 1. Why most code-generated specs fail in AI-driven workflows, 2. The difference between design-first and context-first, 3. How to structure APIs that LLMs can understand and assist with, 4. A demo of the MCP server turning OpenAPI into real-time AI help
Whether you believe in design-first or not, context-first might change how you think about specs—and how you work with AI.
My MCP server is an AI agent
We will demonstrate how to create MCP services that are sophisticated AI agents themselves. They utilize domain-specific and finetuned local AI to perform complex actions for the planning LLM.
We will further discuss how to securely manage and share your MCP services, which makes them accessible for other AI agents within your own network or across the internet.
Empowering Context-Aware APIs with Kong's MCP Server
This session aims to bring together API architects, developers, DevOps engineers, and business leaders to explore how Kong MCP Server is redefining the way we manage and leverage contextual data in modern applications. Whether you're building APIs for cloud-native environments, microservices architectures, or IoT ecosystems, this session will equip you with the knowledge and tools to harness the full potential of context-driven technologies.
Beyond the Spec: Practical MCPs Through LangDB AI Gateway
MCPs (Model Context Protocol) are emerging as a key layer in LLM orchestration—much like HTTP did for the web. But the ecosystem is still rough. The official spec for auth is complex and still evolving. Versioning is largely unsolved, and there are real concerns around prompt injection and untrusted third-party code. In this talk, we’ll unpack these challenges and explore how the AI gateway pattern can help address them. We’ll demonstrate how LangDB provides built-in visibility into MCP interactions, supports tracing and debugging across agent workflows, and share our thinking on what’s needed next in MCP tooling—including some of the features we’re building to help developers use MCPs in production.
Big thanks to ReAPI Kong, Gaia and AWS for supporting our event!