By Dipendra Shekhawat 
Imagine this for a moment.
You say,
“Hey Claude, order me a pizza from Zomato.”
And instead of replying, “I can’t access that website,” Claude actually places the order for you - no plugins, no browser extensions, no manual steps.
It directly talks to Zomato, gets your location, fetches restaurants, selects a pizza, and places the order.
Sounds like magic, right?
Not quite. That’s MCP - Model Context Protocol, and it’s quietly redefining how AI connects with the real world.
Let’s break it down.
🧩 What is MCP (Explained Simply)
MCP (Model Context Protocol) is like a universal translator between your AI model and the outside world.
Think of it this way: your AI (like Claude, GPT, or Gemini) is like a super-smart kid locked inside a room. It knows everything but can’t act outside its four walls.
MCP opens that door.
It lets your AI model connect to tools, databases, APIs, and applications in a standardized and secure way.
Just like Wi-Fi connects your phone to the internet, MCP connects your AI to real-world systems - Zomato, Notion, Jira, Salesforce, SharePoint, and beyond.
So instead of hardcoding hundreds of integrations, you give your AI a universal key to talk to anything, anywhere.
🚀 Why MCP Matters
For years, we’ve seen AI models get smarter in reasoning and conversation — but they’ve been blind to live data.
They couldn’t access:
- Your business systems
- Your APIs
- Your operational tools
We tried using plugins, connectors, and APIs - but each one was different, messy, and hard to maintain.
That’s where MCP changes the game.
It provides AI models with a common, standardized protocol for safely and efficiently communicating with any external system.
With MCP:
- Any AI model can connect to any app
- Developers don’t need to reinvent the wheel
- Everything stays secure and scalable
It’s like giving your AI a universal remote control for the digital world.
🧠 Step-by-Step: Connecting Claude to Zomato via MCP
Let’s see how this works in action: Claude connecting to Zomato.
Here’s how you can set it up yourself:
1️⃣ Download Claude Desktop
Go to claude.com/download and download Claude Desktop for your operating system (Windows or Mac).
Install it and log in with your Anthropic account.
2️⃣ Locate the Config File
Find the configuration file named:
claude_desktop_config.json
You’ll usually find it in your AppData or system folder.
Open it using a text editor.
3️⃣ Add the Zomato MCP Server Configuration
Paste the MCP server details for Zomato (available in the video or documentation) into your config file.
Save it, close Claude Desktop completely, and reopen it.
Now, you should see Zomato MCP Server listed as one of the connected tools.
4️⃣ Authenticate Zomato Access
When you first try to order food, Claude will ask for OAuth authentication to connect to your Zomato account securely.
Once authenticated, the connection is saved, so you won’t need to log in again each time.
5️⃣ Test It Out
Now simply type:
“Claude, order me a pizza.”
Claude will:
- Fetch your saved Zomato addresses
- Suggest nearby restaurants
- Show available menu options
- Let you pick a size or toppings
- Create your cart
- Proceed to checkout (using Pay Later option for safety)
All through natural language conversation - no manual clicks, no web browsing, no coding.
That’s the power of MCP in action.
🤖 The Future of Agentic AI
What you just witnessed isn’t a simple API call.
It’s a glimpse into the next era of AI i.e Agentic AI - where AI doesn’t just talk, it acts.
MCP makes this possible by letting AI models:
- Pull live data from any application
- Take context-aware actions
- Safely interact with systems under your control
It’s open-source, secure, and developer-friendly, meaning anyone can build custom MCP servers to connect their own tools, CRMs, automation systems, or enterprise platforms.
We’re stepping into a future where AI assistants will:
- Book your travel
- Automate your workflows
- Manage your CRM
- Execute your emails
- And yes, even order your pizza
All autonomously, powered by MCP.
👉 Watch the step-by-step video here
⚡ Key Takeaway
The future isn’t just about AI that can think - it’s about AI that can act.
MCP is the bridge between intelligence and action, connecting language models with the real world in a universal, secure, and scalable way.
So the next time you see Claude or ChatGPT doing something surprisingly “human,” remember: It’s not magic. It’s MCP working behind the scenes.
🔗 Final Thoughts
If you found this fascinating, try it yourself. Connect Claude to Zomato and experience what it feels like when your AI really starts doing things for you.
And if you’re an automation enthusiast, MCP isn’t just a tool - it’s a doorway to the future of Agentic Automation.
Because the future isn’t AI alone. It’s AI-connected through MCP.