wellsy.io: our own website, made readable and citable by AI

Wellsy is a web and AI agency based in Dubai, working with businesses. Between August and October 2026, we reworked our own WordPress website: version-controlled code, legal texts rewritten for Wellsy FZCO, two service pages structured for AI assistants, a complete English version and open access for the crawlers of answer engines that cite their sources.

Project at a glance

Sector: Web and AI agency, business services
Location: Dubai, United Arab Emirates
Delivery: August to October 2026
Technologies: WordPress, Voxel theme, Elementor Pro, TranslatePress, The SEO Framework, Schema.org
Languages: French, English
Wellsy offer: Wellsy GEO, Wellsy WEB

Why treat our own website as a client project?

wellsy.io is our agency’s showcase. It is also the first website where we apply Wellsy GEO, our service for visibility in AI answers. We call it our client zero: anything we recommend to clients has to work on our own site first, under the same conditions and with the same checks.

The site already existed, built on WordPress with the Voxel theme and Elementor. We chose to keep that foundation rather than switch platforms. The goal was not to rebuild everything, but to make the site traceable, aligned with the company behind it and readable by generative engines.

What did we change in the site’s foundations?

Since August 2026, all of the site’s code has been under version control. Every change is prepared locally, recorded with its reason, then deployed to the live site. Any change can be traced and reverted without digging through the WordPress admin.

We then rewrote the three legal pages for Wellsy FZCO, a company registered in Dubai. The legal notice, the terms of sale and the privacy policy now apply the law of the United Arab Emirates, reserve our services for business clients and name the courts of Dubai. The privacy policy was rewritten from scratch.

These texts are available in French and English, 147 translated segments in all, and they are linked from the footer in both languages. The Wellsy GEO service has its own terms and conditions, published in August 2026 in French and English.

How did we write the service pages for AI assistants?

The AI visibility page presents Wellsy GEO. We built it by duplicating the template of our website creation page, to keep its layout, then rewrote every text slot. It mentions no amounts: pricing is discussed in a meeting.

Its structure follows the rules we apply for our clients. The AI automation page, which presents Wellsy FLOW, uses the same approach with a six-step project timeline. In October 2026, we grouped both pages under an AI Solutions column in the mega menu, on desktop and mobile, in both languages.

Six of the seven H2 headings phrased as questions.
Under each question, a direct answer paragraph of 40 to 46 words.
An eight-question FAQ marked up as FAQPage, in French and in English.
A search title and description that are translated as well.

Can generative engine crawlers read wellsy.io?

Yes, and we checked it with real requests: ClaudeBot, GPTBot, OAI-SearchBot, PerplexityBot, Googlebot and Bingbot get the same response as a regular browser.

In October 2026, the robots.txt file was opened to four AI crawlers: GPTBot, ClaudeBot, Google-Extended and Applebot-Extended. Other training crawlers remain blocked. The sitemap now lists 114 URLs, 57 in French and 57 in English, and every English page responds correctly.

Frequently asked questions about this project

Why is wellsy.io your client zero?

Because it is the first website where we apply Wellsy GEO. We test page structure, translation, markup and crawler access there before offering them to clients. If something does not work on our own site, it has no place on theirs.

Did you change CMS for this project?

No. The site stays on WordPress, with the Voxel theme, Elementor and TranslatePress. We judged it more useful to version the code, restructure the key pages and open the site to AI crawlers than to rebuild a foundation that was working.

Which AI crawlers can read the site?

Crawlers from engines that cite their sources get through, including ClaudeBot, GPTBot, OAI-SearchBot, PerplexityBot, Googlebot and Bingbot. The robots.txt file also allows Google-Extended and Applebot-Extended. Amazonbot and Bytespider remain blocked, as do several training crawlers unrelated to citation.

Is the site available in English?

Yes. The service pages, legal pages, menus and metadata are available in English under the /en/ prefix. The sitemap lists every page in both languages, so search and answer engines can find the English version without going through the French one.

What about your project?

Working on a similar project? Let’s talk: we will come back to you with a first, concrete opinion on your needs.