Valenys Accessibility
Scheduled axe-core scanning, VPAT and accessibility statement in seven languages, and a client portal to compare runs.
It installs through a WordPress plugin, a Shopify app, an OpenCart module, an npm package, Google Tag Manager, or one line of JavaScript. An audit tells you where you stand today; this is what keeps that answer true.
We ship a visitor widget, and it is genuinely useful: text size, contrast, reduced motion, reading aids, and language options in more than fifty languages including right-to-left scripts, with each visitor’s preferences remembered.
It is not conformity. An overlay on an inaccessible site does not make the underlying content accessible, and selling it as though it did is how this product category earned its reputation. The widget serves the visitor. The remediation, the documents and the monitoring are what serve the obligation.
What the visitor sees
Text size, spacing, alignment, contrast, dyslexia font, reduced motion — in the visitor’s own language, remembered between visits. Useful, and still not conformity: the paragraph above says why, and we would rather say it than let you find out during an audit.
Your pages, on a cadence you set, using our own rule set alongside axe-core — with a 0–100 score and severity across three levels. Journey scans driven by Playwright in a real Chromium browser cover flows that sit behind a login, with screenshots.
Claude, Gemini or GPT — whichever you configure — draft the parts that are language work: alternative text, ARIA labels, link text, simplified wording. A person reviews every draft before it goes into the report, and the report is where it stops: nothing is written to your site, by us or by a model. An unattended system that rewrites pages is how a site ends up with hundreds of confident, wrong image descriptions.
An accessibility statement referencing Directive (EU) 2016/2102 where it applies, a VPAT / Accessibility Conformance Report in HTML and PDF, the scan report, and a conformance badge in three styles — produced in English, Romanian, German, French, Spanish, Italian or Portuguese, chosen per client.
Role-based access, a table of scans with PDF downloads, comparison between runs, re-testing, a suggested fix plan staged over 30, 60 and 90 days, an EN 301 549 checklist, and GDPR export and deletion on request. The portal renders in the client’s document language.
Periodic email reports, webhooks to your other systems, per-client REST API keys, rate limiting, GDPR retention settings, and TOTP two-factor sign-in for administrators. A GitHub Actions workflow runs the scan inside your CI pipeline, so a regression fails the build instead of reaching production.
A WordPress plugin, an embedded Shopify app, an OpenCart module, an npm package for developers, a Google Tag Manager tag, or a single script tag in the page. What the site is built on does not matter to the widget or to the scanner — the scanner reads the rendered page, not your source.
The client portal
Every scan with its score and its documents — report, VPAT, EN 301 549 checklist — in the language set for that client. An agency runs it under its own name; the client never sees ours.
Valenys Accessibility runs on its own site, accessibility.valenys.com, published in six languages — English, Romanian, German, French, Spanish and Italian. It carries the plans, the widget in detail, and a free scan of a single page that returns a WCAG score without an account and without a card.
The compliance documents are produced in seven languages, one more than the site itself: Portuguese is available for the statement and the conformance report.
Agencies get a partner panel: manage your own clients, with branding cascading down to the widget, under your name rather than ours.
No. The widget adjusts presentation in the visitor’s own browser; it changes nothing in your code, and no widget can. Conformity comes from the interface itself, which means someone editing the code — your developers, or us under a separate engagement. We would rather say this plainly than let you discover it during an audit.
No, and neither does anything else in the platform. The models draft suggested wording — alternative text, ARIA labels, simplified phrasing — a person reviews it, and it goes into the report. The platform has no write access to your site and never asks for any.
Seven: English, Romanian, German, French, Spanish, Italian and Portuguese. The language is set per client, and the client portal follows the same setting.
Often enough to catch a regression before it ships to everyone — in practice, aligned with your release cadence, plus a scheduled baseline run. Both are configurable.
Yes. Scan results export as CSV or JSON, an access key lets your own systems read them, and data export and deletion are available on request as standard.
Journey scans use Playwright against a real Chromium browser with an authenticated session, so flows that require signing in can be covered. This is agreed during scoping because it needs credentials and a controlled environment.
Yes. A GitHub Actions workflow runs the scan against a preview deployment and fails the job on a regression, which is how you stop accessibility debt accumulating between audits.
