🗣️ Dysera Speech Therapy

AI-Powered Dysarthria Assessment & Speech Therapy Platform

View Live Demo

📖 Overview

Dysera is an AI-powered speech therapy platform specifically designed to aid in dysarthria assessment and therapy. Utilizing advanced machine learning workflows with FastAPI, Groq LLMs, and a React frontend, Dysera provides patients and therapists with real-time feedback, detailed reports, and personalized AI coaching.

---

🚀 Key Features

• 🎙️ Voice Assessment: Real-time AI detection and analysis of dysarthric speech patterns.
• 🧠 AI Coaching: Personalized feedback driven by Groq to help users correct their speech.
• 📊 Patient & Therapist Dashboards: Comprehensive analytics, progress tracking, and session histories.
• 🔐 Secure Auth & Data: Fully integrated with Supabase for robust user management, roles, and real-time data persistence.
• ⚡ Blazing Fast Backend: Powered by FastAPI for swift processing of ML pipelines and model inference.

---

📸 Screenshots

Landing Page	Login Page

Patient Dashboard	AI Coach

Therapy Exercise	Record & Detect

Patient Report	Profile & Settings

---

🛠️ Tech Stack

Frontend

• Framework: React + Vite
• Styling: Tailwind CSS + Radix UI + Framer Motion
• Routing: React Router

Backend & ML

• Framework: FastAPI (Python)
• AI Models: Groq API / Custom ML models
• Database/Auth: Supabase (PostgreSQL)

---

📁 Project Structure

dysera-speech-therapy/
├── backend/                # FastAPI application, ML models, and API endpoints
├── public/                 # Static assets and screenshots
├── src/
│   ├── components/         # Reusable UI components
│   ├── lib/                # Context providers and utilities
│   ├── pages/              # Main application views and routes
│   ├── App.jsx             # React routing setup
│   └── main.jsx            # React entry point
└── package.json            # Frontend dependencies and dev scripts

---

⚙️ Setup and Installation

Prerequisites

• Node.js & npm
• Python 3.8+
• Supabase Project & Groq API Key

1. Clone the Repository

git clone https://github.com/theadhithyankr/dysera-speech-therapy.git
cd dysera-speech-therapy

2. Install Dependencies

Install both the frontend and backend dependencies. (Assumes Python backend dependencies are listed in backend/requirements.txt)

# Frontend
npm install

# Backend
cd backend
pip install -r requirements.txt

3. Environment Variables

Create a .env file in the backend/ directory:

GROQ_API_KEY=your_groq_api_key
DATABASE_URL=your_supabase_postgresql_connection_string
SECRET_KEY=your_secret_key

4. Run the Development Server

Dysera comes with a concurrently script that starts both the Vite frontend and Uvicorn backend simultaneously.

npm run dev

• Frontend: http://localhost:5173
• Backend API: http://localhost:8000

---

🔗 Links

• Live Demo: https://dysera-speech-therapy-main.vercel.app
• Repository: GitHub

_{Built with ❤️ for better speech therapy.}