AI for the Rest of Us: Vibe Coding & Building MCP Servers with FastMCP

You've heard the buzzwords — AI, LLMs, vectors, embeddings, "MCP" — but you've
never actually sat down and built anything with them. This hands-on day is
for you. No AI experience required, and you don't even need to be a developer.
We'll start at the very beginning: what AI and large language models actually
are, what a "vector" and an "embedding" really mean, and how a vector store
(yes, including SQL Server 2025) fits into the picture. Then we'll get our
hands dirty with "vibe coding" — using AI to build small, working tools live,
breaking things and fixing them as we go. In the afternoon we level up to
FastMCP: what the Model Context Protocol is, how to find and use MCP tools,
and how to build your own so an AI can take real actions — like querying a
database in plain English.
This is a workshop, not a lecture. We introduce a concept, then you build it.
You won't walk out with a polished product — you'll walk out with something
better: a real understanding of how modern AI fits together, and the
confidence to go build your own.
What you'll learn

• What LLMs, vectors, embeddings, and vector stores actually are — in plain
  English
• How to "vibe code" working tools with AI, and what to do when it breaks
• What MCP and FastMCP are, and how to find, use, and build your own MCP tools
• How to get an AI to take real actions (like working with a database) safely
  

What you'll leave with

• A clear mental model of how modern AI systems fit together
• Hands-on experience building (and breaking) things with AI
• Your first MCP tool — and the confidence to keep going

What to bring (prerequisites)

• A laptop (Windows 11+, MacOS M1+) with the workshop environment installed
  ahead of time — we'll send setup instructions before the event so you arrive
  ready to build
• Curiosity. That's it — no prior AI or coding experience needed.