Accessibility Booster – AI-Powered Accessibility Toolkit

AI-powered tools for improved accessibility: summarize, translate, proofread, write, rewrite, and personalize your web experience.

Comment

Creating a truly accessible web for everyone is still a challenge, especially for people with disabilities or language barriers. Accessibility Booster was inspired by the desire to leverage AI to provide advanced, customizable, and easy-to-use accessibility tools directly in the browser.

What it does

Accessibility Booster offers:

  • Summarizer: Condense long texts in different styles (concise, detailed, bullet points).
  • PromptBox: Multimodal prompts (text and image) for explanations, visual descriptions, and creative analysis.
  • Translator: Translates text between 8 languages, with auto-detect and formatting preservation.
  • Proofreader: Checks and corrects grammar, spelling, and clarity, with suggestions and explanations.
  • Writer/Rewriter: Generates or rewrites content for various contexts and audiences.
  • Accessibility Panel: Adjust font, color, contrast, spacing, and navigation. Supports dyslexia-friendly font, dark mode, keyboard navigation, and more.
  • Onboarding: Interactive tour and contextual tooltips for first use.

All settings persist across sessions, making the experience seamless and personalized. Users can add their own Google AI API key easily in settings.

How I built it

Developed using Next.js, React, Tailwind CSS, and integrated with Google Generative AI (Gemini) APIs. The project follows best accessibility practices, with a strong focus on real-world usability.

Challenges faced

  • Handling API model/version limitations and user errors gracefully.
  • Ensuring all tools were accessible with screen readers and keyboard.
  • Providing a straightforward API key onboarding for non-developers.

What I learned

  • How to integrate multiple AI APIs in a single interface.
  • Advanced accessibility design for digital tools.
  • Techniques for managing multimodal AI prompts and processing varied user inputs.

Feedback and reflection

During the hackathon, I learned a lot about practical accessibility design and integrating multimodal AI features. The main challenge was making the interface truly usable for users with diverse needs and connecting robust AI APIs. With additional time, I would refine onboarding and build more adaptive content tools.

Built With

Share this project:

Updates