Lexara Air
For a dyslexic learner a dense exam paper is a wall. Lexara Air takes the PDF and rebuilds it as guided, step-by-step lessons with calmer type, spoken support and your language.
6
Pipeline stages
9+
Indian languages
3
Reading modes
The story
01 · THE PROBLEM
Dense pages punish slower decoding.
Exam content is laid out for fast readers. Spacing, pacing and structure are treated as decoration, when for some learners they decide whether the question can be read at all.
02 · THE PIPELINE
From PDF to lesson in six stages.
Tesseract OCR and GPT-4 Vision read the paper, FastAPI services restructure it into steps, and text-to-speech and translation make it spoken and multilingual. Redis and AWS keep it responsive.
03 · THE EXPERIENCE
Three modes that react to load.
Easy, structured and scaffolded modes adapt layout, wording and pacing, so the same question can meet a learner where they are.
04 · THE LESSON
Accessibility is cognitive, not cosmetic.
Good assistive products do more than restyle text. They change how information is sequenced, spoken and understood.
How it's built
Every box is a real part of the system. The dots are data moving through it.
Built with
- Next.js
- FastAPI
- GPT-4 Vision
- Tesseract OCR
- Text-to-Speech
- Redis
- AWS