I designed my online portfolio in 1 day ✨(with AI...)
π‘This article is a support to present a basic use of AI in software developmentπ‘
ππ» Introduction
In this article, I will show how I imagined, designed, and published my portfolio in a single day, with the help of AI. The goal is not to dive into complex technical details, but to explain in a very concrete way how artificial intelligence can help us today, save time, and simplify certain steps — even without being a developer or an IT expert.
I will conclude with a personal reflection on AI — possibly going against some common assumptions — but which I believe is necessary to put things into perspective and better understand what these tools can (and cannot) really offer.
ππ» Steps
Below are the steps I followed to create my personal portfolio (this is of course my own structured method, but not the only possible one).
Roll up your sleeves and formalize your personal information. I used my old existing portfolio as a starting point and updated it (see presentation below).
Think about the visual form you want to give your portfolio / website. Ideally, the site should match your professional field as well as your personal taste.
Prepare and format the input data that the AI will work with, and define an interface.
Launch the creation of a first version of the site using AI (watch out for technical pitfalls — technologies used, placeholders to get started, etc.) ⇒ recursive step: refine until the site feels right.
Integrate the input data into the site ⇒ ask the AI to rewrite it in the desired style (professional, technical, concise, etc.).
List all images / photos / illustrations needed to showcase your experiences ⇒ ask the AI to refine the images (desired formats: landscape / portrait / square).
Integrate the generated visuals and do a final pass to make sure everything fits properly.
Bonus: multi-language translations using AI (pay attention to the order of steps).
Bonus: SEO, assisted by AI (especially if you’re not sure what SEO actually means).
#️⃣ Step 1
The first step cannot be delegated to artificial intelligence. First of all, you need to roll up your sleeves and take the time to formalize your own information.
This means gathering your most relevant experiences, selecting the projects that truly matter, and structuring everything in a readable way: year, job title, employer, and a clear description of what was done.
In my case, I relied on an old personal blog I had created a few years ago, which I updated and partially reused as the foundation for this new portfolio.
π€ My Old-school portfolio website:
⇒ Everything is presented in a very linear and not very engaging way for an interested reader…
However, I can copy/paste parts of my experiences and reorganize them into a raw text file. I then complete it with additional information coming from my CV, for example.
#️⃣ Step 2: Visual design of the site
The second step is about defining the visual direction of your portfolio or website. It should be consistent with your professional field, but also reflect your personal taste.
If you lack inspiration, you can simply browse existing work by searching keywords such as “portfolio”, “creative portfolio”, or “developer portfolio”. Platforms like:
Behance
Dribbble
Awwwards
ThemeForest
Figma Community
are full of mockups, templates, and well-designed portfolio examples.
However, one essential thing must be kept in mind: even if you like the site, visitors will almost never share the same taste as you. That’s why it’s usually better to stay minimal, readable, and consistent rather than trying to do too much.
⇒ The goal is to describe as precisely as possible what you want to the AI. If you’re not sure which sections really matter, you can also ask the AI for clarification before starting.
#️⃣ Step 3: Formatting data for AI
Before using AI properly, it is essential to do some preparation work. AI is only truly effective if the input information is clear, structured, and easy to interpret — whether it’s text, images, or any other type of data.
I took the time to format my information in a simple and consistent way inside a raw text file. For each experience, I defined a clear structure:
year
job title or mission name
description
list of keywords (in my case, technical / IT keywords)
With this format, AI can retrieve and process the information without ambiguity or errors, which saves a lot of time later.
π‘ AI sometimes makes mistakes — and the most distracted among us won’t even notice them. That’s why, in my opinion, it’s better to prevent issues upfront with clean data.
#️⃣ Step 4: generating the site with AI
Once the foundations are in place, you can launch the creation of a first version of your site using AI. This step allows you to quickly obtain a structure, a layout, and an initial content organization.
π‘However, you must stay cautious about technical choices: some suggested technologies may be outdated, poorly suited, or simply inappropriate. If you’re not comfortable with these topics, don’t hesitate to ask the AI for guidance before going further. If you already have technical habits or skills, sticking to technologies you know will make validation and adjustments much faster.
π‘Using placeholder images and texts at this stage is a great approach: it lets you focus on layout without blocking yourself on final visuals. The definitive images can be added later, once the layout is fixed, and generated or adjusted to the right format and ratio.
Repeat layout refinement cycles until you’re satisfied with the result.
 |
| *this is not the subject but many tools existfor that: I used Antigravity https://antigravity.google |
#️⃣ Step 5: Integrating “raw” data
Once the layout is fixed, you can ask the AI to directly integrate the structured data prepared earlier in your text file. The AI can rewrite this content in the desired tone — professional, technical, concise, or narrative — while respecting the site structure.
All that remains is to check that everything fits correctly and that the final rendering is clear and coherent.
#️⃣ Step 6: Photos and illustrations
The next step is to precisely list all the images, photos, or illustrations needed to showcase your experiences and portfolio sections. Once the list is complete, you can ask the AI to generate or refine all or part of these visuals, respecting the formats required by the site (portrait, landscape, square, etc.).
It’s then important to name files clearly and consistently, and place them in the appropriate image directories. This organization simplifies integration, updates, and future reuse — whether by the AI or by yourself.
 |
| AI can refine / enhance your pictures / diagrams |
Finally, verify image integration and iterate until the result is perfect.
#️⃣ Step 7: Multi-language website
Managing multi-language translations with AI is a huge time-saver, but it must come at the very end of the site creation process. At this stage, the structure, components, and content should already be finalized.
If new texts or components are added later, it’s easy to forget to translate them into all available languages, creating inconsistencies. When AI is focused on creating or modifying a component, it may not automatically generate the corresponding translations — hence the importance of treating localization as a final, global, and controlled step.
 |
| IA can translate your website helping you to target your audience |
#️⃣ Step 8: SEO — making your site more visible online
Finally, for those who have already heard about SEO (search engine optimization), AI can also provide valuable help. It can generate basic elements to improve a site’s visibility, such as page titles, descriptions, keywords, and a structure that search engines can better understand.
Without going into complex technical details, this allows you to have a cleaner site that is better prepared to be found online — even if SEO is not your area of expertise.
ππ» Personal conclusion on using AI
With this small hands-on experiment — which I consider a fairly simple use case for modern AI tools like ChatGPT or Gemini — here is what I can say about AI usage and the reflections it led me to.
π§π»π»
At a time when we see more and more videos about “vibe-coding” (people building apps without knowing programming and relying exclusively on AI), I was also curious to see what I could realistically achieve when creating a website.
→ First point: to summarize briefly, I realized that most “vibe-coder” videos show how to nicely wrap small candies — but rarely how to prepare a full banquet and deal with more complex problems. AI is a very productive tool for many small tasks, but in my experience, only experienced developers can truly leverage it at scale, especially in professional contexts where many non-technical constraints also exist.
→ Second point: Artificial intelligence is a powerful tool, but it is neither the creator nor the decision-maker. Ideas, vision, choices, and overall coherence always come from you. AI can accelerate, assist, and simplify — but it can also make mistakes. It cannot truly put itself in your place. If you don’t understand what you ask for, or if you don’t verify what it produces, you may accept errors — or worse, build something incorrect without even realizing it.
ππ‘
Additionally, based on my research into current AI techniques and tools, it’s important to remember that AI itself consumes a significant amount of resources — both computational and energy-wise. Using AI therefore also implies responsibility on our side.
Clearly preparing project steps, structuring needs, and knowing what we expect before querying AI leads to real benefits. On the other hand, multiplying vague prompts without clear objectives, without verifying results, or without understanding what is produced often leads to little real time savings (or short-term gains followed by long-term losses) and unnecessary resource waste.
A thoughtful and controlled use of AI is therefore key to making it an effective tool rather than just a generator of approximate answers.
ππ»Below is a non-exhaustive list of pros and cons you should be aware of before starting AI-assisted creation:
Advantages of AI | Drawbacks / Warnings
Significant time savings: AI can quickly generate site structures, texts, layout ideas, or basic visuals. | Risk of errors: AI may produce incorrect, approximate, or technically wrong information without obvious signs.
Structuring help: very effective at reorganizing and rewriting prepared data. | Dependency on input quality: poorly structured data leads to poor results.
Accessibility: enables non-technical people to create websites. | Inappropriate technologies: AI may suggest outdated or unsuitable solutions if not guided properly.
Versatility: text, images, translations, SEO — AI can help at nearly every stage. | Lack of global coherence if additions are made progressively without control.
Creative accelerator: helps unlock ideas and test variations quickly. | Illusion of mastery: misunderstanding AI output can lead to complex dead ends.
Quality standardization: helps maintain consistent tone. | Lack of human judgment: AI executes, it does not think critically.
π‘ If you’re motivated, try building your own portfolio with AI and tell me what you think π
π€ Or if you don't have time contact me directly — I’d be happy to help with your projects. It’s always nice to talk to real humans
⇒ https://julien.goenaga.fr
πBonus: Energy impact of AI
π‘ AI & resources: what you should know today
Key facts
Training AI models like ChatGPT requires massive GPU clusters, consuming huge amounts of electricity over short but intense periods.
Daily AI usage (inference) is a continuous cost: each request uses servers, memory, networking, and cooling.
Data centers already account for ~1.5% of global electricity consumption, and AI is one of the main growth drivers.
A single request may seem insignificant, but millions or billions add up.
Exact figures are rarely public — we usually talk in orders of magnitude, not precise values.
Concrete implications
AI is not “free” in energy terms, even if invisible to end users.
Poor usage (useless iterations, vague prompts, redundant requests) unnecessarily increases resource consumption.
Companies constantly optimize cost per request, but growing demand often offsets these gains.
Users risk false productivity: many interactions, little real value.
AI can produce silent errors that are hard to detect without domain knowledge.
Responsible usage best practices
Prepare projects upfront: clear goals, defined steps, structured data.
Limit unnecessary back-and-forth: one precise prompt is better than ten vague ones.
Always verify outputs, especially for technical or critical topics.
Use AI as an assistant, not a decision-maker or source of truth.
Accept that value comes from humans: AI accelerates, but does not replace vision or judgment.
Comments
Post a Comment