About AccessWeave

The problem

People with disabilities face an accumulation of small barriers — a label that can't be read, a form written in dense language, a step that assumes a phone call. Most tools solve one modality (a screen reader reads; OCR extracts). The burden of stitching them together still falls on the disabled person.

Our approach

AccessWeave treats accessibility as an end-to-end task-completion problem. It asks not “what is your diagnosis?” but “what would make this task accessible to you?” Then it compiles the task — parses the source, detects barriers, builds an explainable task graph, applies your Access Passport, and renders validated Access Cards in your preferred format.

What makes it different

• Adapts to the task, not a diagnosis label.
• Transforms experiences instead of only detecting problems.
• Unifies digital and physical accessibility.
• Preserves user agency — consequential actions always require confirmation.
• Works in degraded conditions: a deterministic engine runs offline, with no paid API.
• Measures independence, not engagement.

Built to be accessible itself

AccessWeave targets WCAG 2.2 AA: full keyboard operation, visible focus, skip links, semantic landmarks, no color-only status, text resizing to 300%, reduced motion, high contrast, light/dark themes, and screen-reader announcements. Every page works without JavaScript; JavaScript only enhances it.