MCP — Model Context Protocol — is the AI infrastructure that's quickly becoming the key layer underneath almost every serious AI setup. It's a big part of why AI is shifting from something where you copy and paste from one tab to the other, to something that actually can act on your behalf, and is the foundation that makes agentic AI possible.

Last week in part one of this series, we covered the what MCP is and why it's such a big deal. In part two, we get into what MCP actually looks like in practice, from the easy non-technical entry points (no technical knowledge required!) already built into Claude to automation tools like Zapier to Kyle's own more advanced setup that allows you to have plain-English conversation with your business data in 30 minutes. We cover a range of options so you can find the right starting point for where you are right now, and understand how far you can take it from there.

If you are a founder, operator or small business owners who is tired of manually looking at data across all your different systems and tools, this is the episode for you.

What You Will Learn

The easiest way to get started with MCP with no technical knowledge required
What a more advanced setup looks like using BigQuery and Claude
Why clean data still matters — MCP removes the barrier between you and your data, but it can't fix what is broken underneath
The safety rules that apply to every MCP setup
What multi-agent systems look like next, and why MCP is the infrastructure that makes them possible

Try Granola

If you've ever sat in a meeting, taken what felt like decent notes, and then opened them afterwards and they didn't capture anything, Granola is the tool for you. It runs in the background, captures everything, and turns your notes into something you can actually use. Both Jess and Kyle use it, it plays really nicely with Claude, and it is one of our most highly recommended tools

New users get 100% off their first month using our link: granola.ai?via=early-adoptr

Timestamps:

00:00 What We've Been Up To This Week

04:43 What Is MCP and Why Does It Matter? A Quick Recap

08:58 Why AI Agents Need MCP to Actually Be Useful

11:28 The Easy Wins: MCP Connectors Already Built Into Claude

18:12 Zapier, n8n and Make: The Next Step Up

23:40 The Advanced Setup: Talking to Your Data Warehouse With Claude

26:28 The Problem MCP Solves: Getting Answers Without a Developer

28:47 Asking Your Data Questions in Plain English

32:41 Democratizing Data Analysis for All Businesses

34:33 Garbage In, Garbage Out: Why Clean Data Still Matters

36:31 Having a Real Conversation With Your Data: Memory and Context in Data Conversations

39:35 Pulling From Multiple Systems in a Single Question

42:27 Where MCP Is Heading in the Next 12 Months

47:03 AI News of the Week: What 81,000 Claude Users Actually Want From AI

49:06 AI Gone Wrong: The Importance of Human Oversight

52:34 Wrapping Up for the Week

Get in Touch

Email: hello@earlyadoptr.ai

TikTok: @early_adoptr

Instagram: @early_adoptr

YouTube: @early_adoptr

LinkedIn: @early_adoptr

Resources:

Official Anthropic MCP server list: github.com/anthropics/mcp-servers
GitHub MCP server (what we used): github.com/github/github-mcp-server
BigQuery MCP server options: search Smithery.ai for “bigquery”
Zapier MCP (no-code entry point): zapier.com/mcp
Smithery.ai — browse and discover MCP servers
OWASP MCP Top 10 — security reference: owasp.org/www-project-mcp-top-10
AI News of the Week: https://www.anthropic.com/features/81k-interviews
AI Gone Wrong: https://fortune.com/2026/03/18/ai-coding-risks-amazon-agents-enterprise/
AI Gone Wrong: https://techcrunch.com/2026/03/18/meta-is-having-trouble-with-rogue-ai-agents/

Get in touch with Early Adoptr: hello@earlyadoptr.ai

Follow Us on Socials & Resources:

IG: https://instagram.com/early_adoptr

TikTok: https://tiktok.com/@early_adoptr

YouTube: https://www.youtube.com/@early_adoptr

Substack: https://substack.com/@earlyadoptrpod

Hosted on Acast. See acast.com/privacy for more information.

﻿Try Granola

Try Granola