AI Automation
You've heard what AI can do. Now learn how to make it do things for you. This course is about building, not theory. You'll take AI out of the chat window and wire it into your actual life — connecting it to your files, your calendar, your email, your tools. Every section teaches a new capability by adding it to something real: a personal AI assistant that you build from scratch and keep when you're done. You'll start with the basics — how to talk to AI models through APIs rather than chat boxes, and why the way you structure a prompt changes everything about what you get back. Then you'll give your assistant the ability to act: searching the web, calling APIs, running calculations, choosing the right tool for the job. That's the shift from chatbot to agent. From there, you'll connect the agent to your world through the Model Context Protocol — the standard that lets AI read your files, check your calendar, and access whatever systems matter to you. You'll build a custom integration for something specific to your workflow. Then you'll make it work while you sleep. Using n8n, you'll build visual workflows that trigger automatically — morning briefings, inbox monitoring, document processing. The assistant stops waiting for you to ask and starts anticipating what you need. You'll give it memory by connecting it to your documents through retrieval-augmented generation, optionally run it locally for privacy, and even add voice so you can talk to it instead of typing. And you'll finish by making it reliable — adding guardrails, managing costs, handling errors, and building the kind of trust that lets you hand off real work to an AI system. Two tracks run side by side throughout: a code track using the Anthropic Agent SDK for developers, and a no-code track using n8n for everyone else. Both build the same assistant. Choose your path, or do both. No AI expertise required. Just a willingness to build.
Section 1: The Shift — From Models to Systems
Something changed. AI stopped being a research topic and became a building material. The models are already trained — the question now is what you wire them into. This section is about seeing the new landscape: the layers of a modern AI system, the difference between asking a question and engineering a reliable process, and the API call that turns a language model from a website you visit into a component you control.
Section 2: AI Agents — Reasoning and Acting
A chatbot answers questions. An agent solves problems. The difference is the ability to act — to search for information, call an API, run a calculation, and decide what to do next based on what it finds. This section teaches you how agents work, how to give them tools, and how to build one that actually does useful things without spinning in circles.
Section 3: MCP — The Universal Connector
Your agent can search the web and call APIs. But it doesn't know anything about *you* — your files, your calendar, your notes, your systems. The Model Context Protocol changes that. MCP is a standard way to connect AI to any data source or tool, and it's turning AI from a generic assistant into a personal one. This section teaches you what MCP is, how to find and use existing servers, and how to build your own.
Section 4: No-Code AI Automation
So far, your assistant waits for you to ask. This section changes that. Using n8n — an open-source workflow automation platform — you'll build visual workflows that trigger on their own: a morning briefing that summarises your email, a monitor that flags important documents, a pipeline that processes incoming requests while you're asleep. The assistant becomes proactive.
Section 5: The Open-Source Toolkit
Your assistant has skills and connections, but it doesn't remember what you told it yesterday. It can't find the answer in your meeting notes from last week. And you have to type every interaction. This section fills the gaps: retrieval-augmented generation gives it memory, local models give it privacy, and voice interfaces let you talk to it like a person.
Section 6: From Prototype to Production
Your assistant works. It reasons, acts, reads your data, runs automated workflows, searches your documents, and talks to you. But "works" and "works reliably" are different things. This final section is about the gap between a demo and a system you trust — choosing the right architecture, defending against misuse, controlling costs, handling failures, and building the confidence to hand off real work to an AI.