Duration 15 days approx.
These courses are being updated with artificial intelligence. We try to stay at the front of what AI can do: the tools change, and the syllabus is revised so that what you learn still holds when you publish.
Added to the cart. See the final price.
Aurora teaches you to use artificial intelligence in the design of a website without publishing the first thing the model returns. The tool proposes. The person decides. The course starts from a concrete business and from constraints: colours that already exist, a tone of voice, what must not be invented, which photo is real and which one cannot be used. Without those constraints, the AI fills the page with a generic premises, an interchangeable slogan and a structure that could belong to any trade.
You learn to ask for variants with a precise brief. Not “make me a pretty website”, but a home page with one headline, three services and one action, at a phone width, with an accent colour and without stock beach photographs. Several outputs are compared, and what could not be built, or what contradicts the brand, is discarded. Hierarchy is looked at: if the model has given the same weight to everything, it is reordered. If it has invented an award, a founding year or an address, that is deleted. Generated text is treated as a draft: every fact is checked against what the business can stand behind.
On the visual side, the proposals are used to explore composition, not to replace judgement. A suggested grid is turned into real measurements. A suggested button is checked against the background. A suggested palette is reduced to the colours already decided in Lienzo, if that work exists, or to a few written-down codes. Generated images, if they are used, are marked as such when that is needed, and they are not presented as photos of the premises. A fake photo of a team or of finished work is a trust problem, not a shortcut. A simple photo of your own is preferred to a perfect scene that never happened.
The workflow is organised too. The brief, the chosen variant and the reason for discarding the others are saved, so the process can be repeated when the tool changes. A block that nobody has read is not pasted into the site. There is a review list: true facts, contrast, a single action, text that can be read aloud, and no empty phrases. AI speeds up the proposals. It does not speed up the responsibility of publishing. The course insists on that order because the tools are updated and the craft of reviewing does not expire with them.
In a week variants of one page are produced and screened, and only what has been reviewed is published. In fifteen days the same method is applied to several sections, and a reusable brief guide is left. In a month a site already made with AI help is reviewed, and what was invented, what is generic and what cannot be maintained is corrected. Aurora does not train model specialists, and it does not promise that the tool will design on its own. It trains someone who wants to stay ahead in the use of AI and, at the same time, publish a website that still belongs to the business.
Español
English