← All Projects
Building

Accessibility Automator

An AI-powered tool that scans course slide decks for WCAG accessibility violations — missing alt text, low contrast, broken reading order — and suggests fixes automatically. Built to solve the real deadline pressure Temple faculty face with accessibility compliance.

Python Canvas API Claude AI WCAG PowerPoint
⌥ GitHub – To be announced

The Problem

Temple faculty face hard deadlines for making course materials WCAG-compliant. Manually fixing accessibility issues across dozens of slide decks is tedious, time-consuming, and easy to get wrong.

What We’re Building

A tool that accepts a PowerPoint file (or pulls slides directly from Canvas), scans every slide for common accessibility violations, and uses an LLM to suggest or automatically apply fixes:

  • Missing alt text — AI generates contextual descriptions for images
  • Low contrast — flags color combinations that fail WCAG AA
  • Reading order — detects when screen readers would read slides out of order
  • Missing slide titles — required for screen reader navigation
  • Canvas integration — process an entire course’s materials in one click

Why This Matters

Microsoft’s built-in checker misses most contrast issues and has no AI-powered fix suggestions. No existing tool is integrated into Canvas. This fills a real gap for our faculty.

Get Involved

Interested in contributing? Check out the GitHub repo or email us.