Sajeetharan Sinnathurai

@sajeetharan @kokkisajee

• Principal Product Manager at Microsoft — 13+ years in the industry
• Working on Azure Cosmos DB developer experiences
• GDE, MCT & MVP from Sri Lanka
• Top Stack Overflow contributor & top mentor on ADPList
• Passionate about creating tools that empower developers

AI Is Rewriting the Developer Playbook

The fastest transformation in how software gets built.

The Developer Intent Gap

Developers know what they want. The friction is in how.

What is the Model Context Protocol?

How MCP Works

MCP Servers Developers Love

Plug any of these into your AI editor and start talking to your tools

// .vscode/mcp.json
{ "servers": { "my-server": { "type": "stdio", "command": "npx", "args": ["-y", "@package/mcp-server"] } } }

MCP Beyond the IDE

MCP isn’t just for local dev — it powers production AI agents too

Before & After MCP

MCP is Everywhere Now

Growing adoption across AI hosts and tool providers

Build Your Own MCP Server

It's simpler than you think — wrap any API, CLI, or service

import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";

const server = new McpServer({ name: "my-server" });

server.tool(
  "search_issues",           // tool name
  "Find issues by keyword",  // description for the AI
  {                           // input schema (Zod)
    repo:    z.string().describe("owner/repo"),
    query:   z.string().describe("Search keyword"),
  },
  async ({ repo, query }) => {  // handler
    const issues = await github.searchIssues(repo, query);
    return {
      content: [{
        type: "text",
        text: JSON.stringify(issues, null, 2)
      }]
    };
  }
);

Get Started