Prerequisites

This course is hands-on. To get the most out of it, make sure you have the following in place before you start.

Knowledge

• TypeScript fundamentals — types, interfaces, generics at a basic level.
• Modern JavaScript — async/await, Promises, ES modules (import/export).
• The terminal — running commands, setting environment variables.

You do not need prior experience with machine learning or LLM internals. The course builds the agent concepts from first principles.

Tooling

Tool	Version	Notes
Node.js	20 LTS or newer	The SDK and tutorials target modern Node.
Yarn	1.22.x	The repo pins Yarn via packageManager.
Git	any recent	To clone the repo and follow along.
An editor	VS Code recommended	TypeScript IntelliSense helps a lot.

Accounts & keys

• An OpenAI API key with access to the models used in the lessons (e.g. gpt-4o-mini and text-embedding-3-small). You'll place it in a git-ignored .env file — see Getting Started.

Cost note: the tutorials use small, inexpensive models and short prompts, but every run makes real API calls. The tests, by contrast, never hit the network and require no key.

Verify your setup

node --version    # >= 20
yarn --version    # 1.22.x

Once these check out, continue to Getting Started.