Images were extracted from EPUB and layout files in bulk. Invicta generated contextual alt text using AI models trained on educational content. Subject matter experts then reviewed and refined the descriptions, ensuring discipline accuracy for diagrams, charts, and instructional graphics.
This replaced inconsistent manual authoring and reduced the need for author intervention.
Files were processed through PREP to automate the majority of structural remediation, including tagging, reading order, semantic structure, tables, lists, and navigation elements.
PREP automated approximately 90% of remediation tasks. Accessibility specialists addressed complex remaining items and verified compliance. Files were then exported back into EPUB and PDF formats without disrupting layout integrity.
Rather than requiring the publisher to change vendors, Continual Engine standardized accessibility across them.
Workflow: composition vendor → automated processing → expert review → validated accessible output
Automated reporting identified missing alt text, structural issues, and compliance gaps. This allowed accessibility to occur during production instead of after production.
"*" indicates required fields