Building With AI

You can write some code. You have used a chat model. Now you want to put one inside something real — a support triage tool, a search box that answers questions, a feature your users hit a thousand times a day — and you have found that the gap between a good demo and a working feature is wide. This course closes that gap. Nine lessons: the HTTP call underneath everything and what it costs, how system prompts really behave, structured output and its limits, tool use and who is actually running the code, retrieval done properly, chunking and embeddings with real numbers, what an "agent" is once the marketing is removed, how to evaluate your own feature instead of guessing, and the failure modes to design for before launch. Every lesson has code you can run. The course is honest about what does not work: schemas do not make answers true, retrieval does not stop hallucination, and most agent demos fail for reasons no bigger model will fix. It does not cover training or fine-tuning models. This is about building with them.

1. 1The call underneath everything6 minThe API is stateless: every turn you pay to resend the entire conversation.
2. 2System prompts, user turns, and who the model listens to6 minA system prompt is the strongest thing you say, not a rule the model cannot break.
3. 3Structured output, and what a schema cannot promise7 minSchemas guarantee shape, never truth: a perfectly valid object can be perfectly wrong.
4. 4Tool use: the model never runs anything8 minThe model can only ask for a tool call; your code runs it, so your code must authorise it.
5. 5Retrieval: what it fixes and what it does not8 minRetrieval decides what the model can know; when RAG fails, it is usually search that failed.
6. 6Chunking and embeddings in practice8 minEmbeddings rank what a chunk is about, not whether it answers you, so rerank before trusting the top hit.
7. 7What an agent actually is, and why the demos fail8 minAn agent is a loop with a step budget; reliability multiplies, so short chains beat clever ones.
8. 8Evaluating your own AI feature8 minA prompt change is a code change; if you cannot re-run fifty saved cases, you are guessing.
9. 9The failure modes to design for9 minText you did not write is data; it must never be able to trigger a privileged action.