Building with AI: APIs and the Ecosystem
You already know how to talk to an AI model in a chat window. This lesson shows you how real software talks to a model — through a doorway called an API — and introduces the whole family of tools that developers plug together to build AI apps. By the end you could sketch the design of your own AI product.
- Explain in plain words what an API is and what happens when your program “calls” a model.
- Read a real JSON request and response, line by line.
- Understand API keys, why you must keep them secret, and how token-based pricing works — including doing a cost calculation by hand.
- Describe how a system prompt secretly gives a model its role and rules.
- Name the main parts of the modern AI ecosystem: model providers, vector databases, frameworks, and no-code builders.
- Design a small study-helper app on paper, tracing how data flows from a student’s question to a cited answer.
- List the careers and skills that this ecosystem creates.
When you use a chatbot website, a human types a message and reads the reply. But most AI you meet in the real world is not a chat window at all — it is hidden inside other software. The translate button in a messenger app, the “summarize this email” feature, the customer-support bot on a shop’s website: none of these have a person copy-pasting into a chatbot. Instead, the program itself sends the question to a model and gets the answer back automatically.
The doorway that lets one program ask another program for something is called an API. Learning to use an AI model’s API is the single skill that turns you from an AI user into an AI builder. Everything else in this lesson — keys, tokens, system prompts, databases, frameworks — are the tools that hang off that one idea.