Large language models
are not simple to learn.
This course is.
Most people who try to understand LLMs get lost in jargon within the first hour. This program starts where you actually are and builds from there — steadily, without shortcuts.
What the months inside this course actually feel like
Progress here is not linear — it comes in phases, each with a distinct texture and a different kind of challenge.
Confusion that resolves itself
The first weeks feel dense. Tokenisation, attention mechanisms, embedding spaces — the vocabulary is unfamiliar. Most learners describe week three as the moment things start clicking into place.
Confidence through repetition
Mid-course is where the ideas compound. You stop memorising and start reasoning. Exercises move from guided walkthroughs to open-ended problems with no single right answer.
Judgement over instruction
By the final module, the questions change. Instead of asking what something is, you start asking whether it is the right tool. That shift — from knowledge to judgement — is what the course is actually building toward.
Where most people are when they arrive — and what changes
The typical learner joining this program can describe what ChatGPT does but cannot explain why it behaves the way it does. That gap — between using a tool and understanding it — is exactly what this course addresses.
Finishing the program does not make you a researcher. It makes you someone who can read a model card, evaluate a fine-tuning approach, and have a substantive conversation with an ML engineer without nodding along blankly.
Read the full program outlineWhat joining this actually involves
There is no shortcut version of this material. The program asks for consistent effort — not intensity, but regularity. Most people who finish are those who treat it like a commitment rather than a resource.
Ask a question before enrolling