Your network needs smart locks​

Day 6: You built the roads (route tables)

Today: We add the smart locks (Security Groups)

Here's the problem:

Your routing works, traffic can flow.

But there's NO security layer yet.

Right now:

• Anyone can try to connect to your containers
• No firewall rules
• No access control

That's a security issue.

Solution: Security Groups

Your gates need directions​

Day 5: You built the back gate (NAT Gateway)

Today: We build the roads that connect everything

Here's the problem:

You have:

✅ Neighborhood front gate (Internet Gateway)
✅ Back gate (NAT Gateway)
✅ Front yards (public subnets)
✅ Back yards (private subnets)

But nothing is connected yet.

Your back yard needs an exit​

Day 4: You divided your VPC into front yards (public) and back yards (private)

Today: We build the back gate (NAT Gateway)

Your neighborhood needs houses​

Day 1: Your AI agent's first phone call Day 1 ↗
Day 2: Give your AI agent a real-world mission Day 2 ↗
Day 3: You claimed your territory (VPC) Day 3 ↗

Today: We'll build our neighborhood with subnets

Your AI agent needs a home​

Days 1-2: Your AI caller agent ran on your laptop.

Today: We're moving it to the cloud

Give your AI agent a real mission​

Yesterday, your AI agent called you and you had a basic conversation.

Today? We're giving it a real job: booking restaurant reservations.

By the end of today, your AI agent will:

✅ Call a phone number
✅ Act as your personal assistant
✅ Request a dinner reservation
✅ Handle follow-up questions
✅ Sound professional and polite

And you'll test it by answering the phone and roleplaying as the restaurant.

This is where it gets fun.

Let your AI make real phone calls​

Your AI agent is trapped.

It can write emails. Answer questions. Generate code.

But it is stuck in your chat text box. It can't pick up a phone and call someone.

But what if it could?

What if you could say:

"Hey ChatGPT, call the restaurant and book me a table at 7pm"

or

"Hey ChatGPT, call the doctor's office and reschedule my appointment"

What if your AI agent could exist in the REAL world and not just in your browser?

That's what we're building.

Why I built this​

I kept realizing after scrolling that I'd missed Linkedin posts from people I actually wanted to engage with.

Fix: highlight "inner circle" directly in the feed:

• No backend
• No login
• Instant cues

MCP (Model Context Protocol) is a new open protocol designed to standardize how applications provide context to Large Language Models (LLMs).

Think of MCP like a USB-C port but for AI agents: it offers a uniform method for connecting AI systems to various tools and data sources.

This post breaks down MCP, clearly explaining its value, architecture, and how it differs from traditional APIs.

LlamaIndex makes it easy to build AI-powered search, but if you're working with multimodal embeddings (text + images), like the AWS Titan multimodal model, you'll notice it's not natively supported.

For e-commerce search, I need embeddings that capture both product descriptions and images to generate more accurate search results.

This guide will show you how to override LlamaIndex's default embedder to use AWS Titan Multimodal.