LexiBridge.AI AI-Powered Dyslexia Support Ecosystem

Empowering dyslexic learners with AI-driven reading support, gamified learning, and real-time analytics

Comment

Inspiration

Dyslexia affects over 700 million learners globally, yet support systems remain fragmented. Some tools help with reading, others with phonics, a few with assessments—but none provide a unified, end-to-end journey for a child who struggles with reading.

After building two separate prototypes—LexiBridge (web-based reader) and DysLex Helper (gamified mobile learning)—I realized both solved different parts of the same problem.

The insight was simple:

“What if a complete AI ecosystem existed where a child could learn, play, read, practice, track progress, and get early screening—all in one place?”

LexiBridge.AI was inspired by:

  • children who feel discouraged while reading
  • parents who struggle to identify early signs
  • teachers who need deeper insights
  • the belief that every child deserves to read with confidence

What it does

LexiBridge.AI is an AI-powered dyslexia support ecosystem that brings together multiple assistive tools:

1. AI Early Screening

  • Camera-based eye-pattern tracking
  • Speech-to-text reading mismatch detection
  • Error pattern profiling using ML
  • AI-generated Dyslexia Risk Index

2. Adaptive Reading Assistance

  • Real-time text simplification
  • AI narration
  • Syllable breakdown & phonetic cues
  • Fluency and accuracy analytics

3. Gamified Literacy Learning

  • Phonics builder
  • Rhyme matching
  • Sound identification
  • XP, streaks, badges, rewards
  • Difficulty adjusts based on AI

4. AR ReadAssist Mode

  • OCR-based real-time text capture
  • Live simplified overlay
  • Pronunciation hints
  • Highlight path guidance

5. Parent/Teacher Dashboard

  • Reading progress graphs
  • Error-pattern heatmaps
  • Consistency tracking
  • Weekly AI recommendations

6. Offline Support

  • Games
  • Local reading modules
  • On-device AI inference

How we built it

LexiBridge.AI combines features from both earlier projects into a single unified architecture.

Tech Stack

Frontend: Next.js, React, Tailwind, shadcn, Flutter
Backend: Firebase, Firestore, Cloud Functions, Node.js
AI/ML:

  • Gemini (text simplification)
  • Google STT/TTS
  • OCR (Vision API / MLKit)
  • Error-pattern clustering ML
    AR: ARCore + WebAR
    Offline: SQLite + IndexedDB
    Analytics: Firebase Analytics

System Architecture Summary

  • Mobile/Web apps → API gateway
  • Gateway → AI Services (LLM, OCR, STT, TTS)
  • Gateway → Firestore Database
  • ML models → Risk Index, reinforcement learning for adaptive difficulty
  • Offline-first modules → Sync when connected

Design Process

Created using Figma + FigJam:

  • User flow diagram
  • Wireframe mockups
  • Architecture diagram

Challenges we ran into

1. Merging two very different app concepts

Combining a web reader with a gamified mobile app required rethinking navigation, structure, and data flow.

2. Designing child-friendly yet adult-usable UI

The interface had to work for:

  • children (simple, colorful)
  • parents (informational)
  • teachers (data-heavy)
  • specialists (detailed analytics)

3. Building accurate low-latency AI screening

Eye-pattern estimation, speech mismatch detection, and error profiling were complex and required experimentation.

4. Implementing offline-first design

Ensuring stable sync logic between local storage and Firestore was challenging.

5. Accessibility requirements

Text-to-speech, spacing, contrast modes, and dyslexia-friendly fonts had to work across devices.

Accomplishments that we're proud of

  • Developed an end-to-end AI ecosystem for dyslexia support
  • Built AI screening with speech mismatch + pattern analysis
  • Designed gamified learning proven to boost literacy engagement
  • Created AR overlays for reading real books
  • Achieved offline compatibility for underserved regions
  • Built clean Figma wireframes, flows, and architecture

But most importantly:

It helps dyslexic learners build confidence — one word at a time.

What we learned

1. Empathy-first AI design

Assistive technology must adapt to the user, not vice versa.

2. Multi-modal AI dramatically enhances learning

Vision + Speech + Language models = powerful literacy support.

3. Gamification increases retention

Kids learn better when learning feels like play.

4. Accessibility isn't optional

Small changes—contrast, spacing, narration—make major differences.

5. Real-world testing is essential

Iterating with real children/parents revealed insights no spec sheet can capture.

What's next for LexiBridge.AI — AI-Powered Dyslexia Support Ecosystem

1. Launching pilot programs

Partnering with schools, NGOs, and learning centers.

2. Adding more AI assistance

  • Emotion-aware reading coach
  • Personalized reading path
  • Conversational AI tutor

3. Expanding educator dashboards

Full classroom analytics for teachers.

4. Multi-language support

Hindi, Spanish, Arabic, French, more.

5. Launching mobile app on Android & iOS

With enhanced offline-first features.

6. Scaling globally

Bringing affordable, AI-powered dyslexia support to every learner.

Built With

Share this project:

Updates