I founded AIFirstPartner and I have shipped 12 production AI systems for clients in government, legal, and enterprise. The consulting practice is where most of the MCP work happens in my week.
Alongside that, I am fractional CTO at Speculo, a cyber risk software company. My day role is managing cloud and data platform engineering for a large Australian government department supporting 20+ agencies across the portfolio, backed by 15+ years in enterprise technology. The three roles feed each other: what I learn deploying AI for consulting clients lands in the platform, and what the platform needs pushes the consulting work to stay practical.
I publish what I am building week by week at 150days.aifirstpartner.com. It is the unfiltered version: what worked, what did not, and what I changed the next day.
You might also know two tools I build in the open:
The session analyser for AI-assisted development. Upload a Claude Code transcript and get scored across 8 fluency dimensions, with actionable recommendations.
A VSCode extension that monitors your Claude Code sessions in real time: live health scores, anti-pattern alerts, and per-model cost tracking, right inside your editor.
Building those tools, I kept running into MCP as the wiring problem underneath agentic coding. Every team adding a new integration was writing the same tool-schema plumbing. The ones who skipped straight to MCP servers shipped faster and had less to maintain six months later. That is what this course is about.
Model Context Protocol (MCP) is one of the most important shifts in how developers integrate Claude with external services. Most introductions either skim the surface or drop you into reference docs without context.
I built this course as the hands-on path I wished existed when I was learning MCP myself: build a real CLI chatbot end-to-end, see every primitive (tools, resources, prompts) used in a working project, and finish with a clear mental model for choosing between them.
By the end you'll have a working MCP server and client in Python, debugged with the official MCP Inspector, and the decision framework to know when to reach for tools, when to use resources, and when prompts are the better answer.
Think of it as a learning loop:
Learn the method. Introduction to MCP teaches you how MCP servers and clients work and gives you a working Python project to build on.
Practice in real sessions. AI IQ coaches you in real time as you work with Claude Code, catching anti-patterns before they compound.
Review and improve. TranscriptIQ analyses your completed sessions and scores your fluency across 8 dimensions.
Each tool reinforces the others. The course gives you the knowledge, AI IQ gives you real-time feedback, and TranscriptIQ gives you the retrospective view.
Questions, feedback, or just want to say hi. I would love to hear from you.
All tools built by AI First Partner