The year is 2025 and you are manually remediating a PDF. All in all not too bad... You are making progress, click by click, and rest easy knowing that good things take time. That doesn't mean you are not open to looking over the fence. You've heard it all before: One-click automation, accessibility in a breeze. But worry, this is not that that. Instead, this workshop provides an overview of the existing conversion options from PDF to EPUB, highlighting the advances and limitations of Vision Language Models for automatically segmenting image-based documents into EPUB-ready semantic markup. It then introduces AI-assisted tooling and workflows for remediating EPUB source formats, focusing on leveraging VS Code & associated Co-Pilot Pro access as part of complimentary GitHub Education for Teachers access. By becoming familiar with EPUB-ready tooling to complement existing remediation and conversion workflows, attendees can expect to offer more tailored services for students with low-incidence disabilities.

Key Takeaways 

• Compare automation potential in Alternate Media workflows across source formats (PDF vs EPUB)

• Explore capabilities and limitations of vision-language models for document conversion

• Learn AI-assisted EPUB remediation workflows using complimentary GitHub Education tools