About AccessAI

Information shouldn't require a translator, a lawyer, or a medical degree.

AccessAI was built for the CTRL-V Hackathon around one theme: using AI to remove barriers to education, healthcare, services, and opportunity.

The problem

Every day, people face documents they can't fully understand: a hospital consent form full of clinical terms, a government benefits letter in dense legal language, a lease written for lawyers instead of tenants. For someone with limited literacy, a language barrier, or a disability, these documents aren't just inconvenient — they're a genuine barrier to healthcare, housing, education, and opportunity.

The approach

AccessAI turns a photo of any document into something anyone can understand. On-device OCR extracts the text, an AI model rewrites it in plain language, and built-in translation and text-to-speech make sure the result reaches people however they read best — visually, aloud, or in their own language.

Who it's for

Immigrants navigating paperwork in a new language, older adults managing medical letters, students decoding financial aid forms, and anyone who has ever stared at a form and thought, "I have no idea what this means."

Built with

  • HTML5, CSS3, and vanilla JavaScript — no framework overhead
  • Tesseract.js for in-browser optical character recognition
  • GROQ API for explanation, summarization, and translation
  • Web Speech API for text-to-speech
  • LocalStorage for private, on-device history