NOTE: This is a BETA TEST trail of a new class using Zoom. This class will be taught in person when we start classes in Baltimore MD. Beta tests are done using Zoom, but regular classes will be in person only.
(You will need a Zoom account to enter the class)
This class is going to teach you the basics of MCP, Model Content Protocol, for your AI projects. MCP is like a REST API for AI Models. You make a request, your model determines the best tool for answering your question, and then that is used for your query.
Tools can be anything you can put into a Python function. So you can give your AI Access to System Tools such as Ping, or Top or pull in data from web scraping or API's.
Not only can MCP be used for your own projects written in Python, but you can also connect Anthropic Claude and such to your server to provide your own information to have their AI solve your problems.
AI is "Just a Stack", and MCP gives your AI projects custom tools and resources for supporting your requests.
This class will go over:
- What is MCP
- What are Tools
- What are Resources
- Model Selection for MCP
- Writing Server and Client Scripts
- Building a Deployable MCP Web App with Bottle
- Connecting your MCP to Claude Desktop
This class will break down the concepts of what is going on so that even no technical people should be able to have a better idea of how "AI" actually works. Example code will be written in Python and we will step through the code blocks for technical people to follow along.
Workbook and Code Samples will be posted to GitHub before class.
