In this episode of Leap to Scale, Justin Davis and Greg Ross-Munro break down Model Context Protocol (MCP) and explain why it is such a powerful shift in how AI interacts with real business systems.

Starting with a simple explanation of APIs, they explore how MCP allows AI to securely connect to tools like CRMs, time tracking, and project management software so teams can ask questions, automate workflows, and take action using natural language. If you want a practical look at how AI can move beyond chat and become a true partner in scaling your business, this episode is a must-listen.

Episode notes

  • What MCP is and why it exists — An accessible explanation of Model Context Protocol, how it was created, and why it’s a major step forward in letting AI interact with real-world systems instead of just generating text.
  • APIs vs. MCP — A plain-English breakdown of APIs and how MCP builds on them to give AI a standardized way to securely connect to tools, data, and software.
  • Turning AI into a true teammate — How MCP allows AI to read from and write to systems like CRMs, project management tools, time tracking, and invoicing platforms.
  • Real examples from a service business — Using MCP with tools like Jira and Harvest to analyze project data, answer ad-hoc questions, and eliminate manual reporting.
  • The power (and risks) of AI taking action — What happens when AI can book meetings, send emails, and trigger workflows, and why human-in-the-loop design still matters.
  • Why MCP points to the future of work — A look at how MCP could change client interactions, internal operations, and how businesses expose data through AI instead of dashboards.

Episode Transcript

What Is MCP (Model Context Protocol) and Why It Matters

Justin Davis:
Hey, it’s Friday. Welcome in. This is Leap to Scale, the podcast where we help service companies figure out how to grow past their headcount and scale beyond the margin trap. We focus on using products and product thinking to do that. My name is Justin Davis.

Greg Ross-Munro:
I’m Greg Ross-Munro.

Justin Davis:
Greg, how are you doing? It’s been a long week, hasn’t it?

Greg Ross-Munro:
It was a long week on Monday, and today is Friday. I’ve got a flight tomorrow morning, an event in Fort Lauderdale, a flight back the next day, a car appointment, a Hanukkah festival where my kid is singing, and my parents’ 50th wedding anniversary party. Not much of a weekend. Maybe this long episode is just delaying the inevitable.

Justin Davis:
Since we both have laborious weekends ahead, let’s talk about something fun. We’re recording this in early December 2025, and about ten days ago we passed the one-year anniversary of something called MCP.

Greg Ross-Munro:
That is wild. It’s difficult to think about because people listening might have no idea what we’re talking about, but for you and me, MCP is almost the lens through which we see systems and AI now.

Justin Davis:
Totally agree. So let’s start at the beginning. Before we define MCP, should we explain what an API is?

What Is an API?

Greg Ross-Munro:
I’ll take a shot at that. Computers and systems need ways to talk to each other, whether that’s a printer on your desk, a robot arm across the world, or a mobile app storing user data. You can’t just send raw code to these systems, so we use APIs—application programming interfaces.

APIs define a safe, structured way to send instructions and data. Every mobile app, website, and system you use relies on APIs to communicate. The entire internet runs on them.

Justin Davis:
That’s a great explanation. APIs are basically the connective tissue of the internet.

What Is MCP?

Justin Davis:
MCP stands for Model Context Protocol. It was created by Anthropic, the company behind Claude, as a way for AI models to interact with the outside world in a standardized way.

AI models are powerful, but traditionally they’ve been limited to what they already know. MCP changes that by allowing models to securely connect to live systems and data sources. It’s now broadly supported across AI vendors and often gets called the “USB-C for AI,” though that comparison misses the real point.

The real power of MCP is that you can connect AI to systems with almost no code. You don’t need to be a developer to wire AI into your tools and data in real time.

Real-World Example: Jira + Harvest

Justin Davis:
We use Harvest for time tracking and Jira for project management. With MCP servers for both, we can ask questions like:

  • What were the last 10 time entries for this project?
  • What is the average number of hours per task over time?

These aren’t dashboards we built. We just ask the AI, and it pulls the data directly from the systems.

Greg Ross-Munro:
So instead of guessing, the AI can actually check your systems.

Justin Davis:
Exactly.

When AI Can Take Action

Greg Ross-Munro:
If AI can read from systems, it can also act. It could check your calendar, look at your CRM, book a lunch meeting, and email a client.

That’s powerful—and also terrifying if it goes wrong.

Justin Davis:
Which is why human-in-the-loop design matters. As you build longer automation chains, you need intentional checkpoints. MCP enables action, but responsibility still belongs to humans.

Talking to Your Data

Greg Ross-Munro:
One of the most exciting parts is letting non-technical people talk to their own data. You don’t need to be a data scientist anymore.

You can ask questions like:

  • What are our busiest days?
  • When should we staff up?
  • What’s happening in our pipeline?

Justin Davis:
That’s where MCP really shines. Start by connecting it to your accounting, CRM, ticketing, or CMS systems and just ask questions. It immediately shows you the value.

The Future of MCP

Greg Ross-Munro:
Do you think every business will eventually have something like an MCP?

Justin Davis:
I do. Maybe not everyone, but I absolutely see a future where clients interact with companies through AI instead of dashboards.

Imagine clients asking:

  • How many hours have you spent on my project this month?
  • When is the next delivery scheduled?

That’s coming.

Greg Ross-Munro:
That would fundamentally change how companies interface with customers.

Final Thoughts

Justin Davis:
Security is still evolving, but it will be solved. MCP is too important not to be.

Greg Ross-Munro:
If you want to experiment, Google “MCP server directory,” install a few, and start playing. Or just ask your AI how to do it.

If you come up with something wild, give us a call.

Justin Davis:
Greg, enjoy your chaotic weekend.

Greg Ross-Munro:
I’m going to drink some wine with friends.

Justin Davis:
Honestly, same.

Greg Ross-Munro:
Why doesn’t the liquor store have an MCP server yet?

Justin Davis:
That feels inevitable.

Comments are closed.