Skip to main content
A free EquitableDocs tool

AltBridge

Subject experts describe images. Remediators apply the descriptions.

AltBridge connects accessibility remediators with subject-matter experts so the people who know the content can describe the images, and the people who know the standards can apply those descriptions to the document. The expert opens a private link, sees each figure beside the current alt text and the AI draft, and writes a short note. The remediator collects every response in one CSV.

01

Why this exists

Remediators write accessibility metadata for documents they often did not author. When a science textbook arrives for remediation, the remediator can fix tag structure, reading order, and form fields without help. They cannot reliably describe a thermometer reading, a labelled cell diagram, or a colour-coded equation without input from the subject expert.

Today that input is gathered by email, with one Word file going back and forth, screenshots pasted inline, and notes lost between threads. AltBridge replaces that loop with a single review interface. The expert sees each figure beside the current alt text and a draft description, picks a verdict, and writes a short note. The remediator receives all responses in one structured CSV.

02

Who uses it

Three groups of people, working in turn, with no direct contact required between them.

Subject-matter experts review figures from a book they know.

Faculty, course authors, librarians, content owners. No accessibility training. They open a private link in a browser, accept the AI verdict where it is correct, and write a short note where it is wrong.

Remediators send the link, collect the responses, apply them to the document.

In-house accessibility teams, vendor remediators, university disability services. They get one CSV with every figure, every verdict, and every note.

Universities and publishers route batches of titles to the right experts.

Institutional accounts that assemble batches of titles, route each to the appropriate expert, and track turnaround. Available through the EquitableDocs partnership track.

03

How it works

Five steps. No login for the expert. Each upload gets its own private link.

  1. 1

    The remediator runs the document through the pipeline.

    EquitableDocs extracts every figure, drafts a candidate description with AI, and produces a review session.

  2. 2

    The remediator sends the session link to the subject expert.

    No login. The link itself is the credential. Save it like a password.

  3. 3

    The expert opens the link and reviews each figure.

    Each figure shows the rendered page beside the current alt text, the AI draft, and a feedback box.

  4. 4

    The expert keeps, overrides, or annotates.

    Keep the AI verdict where correct. Override where wrong. Add a short note where the alt text needs specific guidance.

  5. 5

    The remediator gets one CSV.

    Every figure, every verdict, every note. Apply it to the document with the standard remediation tools.

Upload a document and start a review

PDF or Word. Each upload gets its own private link. Save the link if you want to come back to your work later.

No file selected

Up to 50 MB. Tagged PDFs work best for the PDF path. For Word files, the tool reads inline images and the surrounding paragraphs as caption context.

Use this to recognise the session later if you upload more than one.

Helps the AI match the alt text to the source. A medical journal needs clinical precision. A Grade 1 textbook needs plain language. Leave blank if you are not sure.

The email of the student or contact who needs the accessible version. After review the file will show as ready to send to this address.

04

Privacy and limits

What we do with your file, and what the tool will not accept.