Skip to content

README.md

Latest commit

 

History

History
175 lines (126 loc) · 3.89 KB

README.md

File metadata and controls

175 lines (126 loc) · 3.89 KB
Raw

🚀 lexi.dev — AI-Powered Sentiment Analysis Platform

lexi.dev is a full-stack, AI-powered sentiment analysis platform built with a modern glassmorphic frontend, a scalable Django REST backend, and a machine learning–driven sentiment classification service. The project is designed to analyze textual data in real time and extract emotions, sentiment polarity, and confidence scores with high accuracy.

This repository follows a monorepo architecture containing:

  • A modern frontend for user interaction
  • A robust backend API
  • A machine learning service for sentiment inference

🧠 Key Features

  • ✨ Modern glassmorphic UI with animated gradients
  • ⚡ Real-time sentiment analysis using AI
  • 🌍 Language detection and translation
  • 📊 Sentiment classification with confidence score
  • 🔗 REST API powered by Django
  • 🧠 Dedicated ML service for inference
  • 📱 Fully responsive across all devices
  • 🛠 Monorepo architecture for clean separation of concerns

🏗 Project Architecture

lexi.dev/
│
├── frontend/ # Vite + React + Tailwind CSS (UI)
├── backend/ # Django REST API (Business logic)
├── ml-service/ # Machine Learning sentiment model
├── .gitignore
└── README.md

🎨 Frontend Tech Stack

  • Vite
  • React.js
  • Tailwind CSS
  • React Icons
  • Glassmorphism UI Design
  • Animated element effects
  • Responsive layout system

Frontend Capabilities

  • Hero section with animated gradients
  • Sticky glassmorphic navbar
  • About section with technology logo loop
  • Interactive sentiment analyzer UI
  • Modern footer with GitHub integration

⚙ Backend Tech Stack

  • Django
  • Django REST Framework (DRF)
  • API-based architecture
  • CORS configuration
  • Text preprocessing & routing
  • Translation & language detection pipeline

The backend serves as the bridge between the frontend UI and machine learning model, exposing clean REST endpoints.

🤖 Machine Learning Stack

  • Python
  • Scikit-learn / Deep Learning models
  • NLP preprocessing
  • Tokenization & vectorization
  • Confidence scoring
  • Saved model inference system

The ML service is responsible for:

  • Sentiment detection (Positive / Negative / Neutral)
  • Confidence score generation
  • Language normalization for better accuracy

🔌 API Workflow

Frontend → Django REST API → ML Service → Sentiment Result → Frontend UI

🔍 Sentiment Output Format

{
  "language": "en",
  "translated": "This product is amazing!",
  "sentiment": "Positive",
  "confidence": "0.94"
}

🚀 How to Run Locally 1️⃣ Clone the Repository

git clone https://github.com/ftashmit/lexi.dev.git
cd lexi.dev

2️⃣ Run the Frontend

cd frontend
npm install
npm run dev

3️⃣ Run the Backend

cd backend
pip install -r requirements.txt
python manage.py runserver 8001

4️⃣ Run the ML Service

cd ml-service
uvicorn app:app --reload --port 8000

🌐 Future Deployment Strategy

  • Frontend → Vercel

  • Backend → AWS / Railway / Render

  • ML Service → Dedicated cloud VM or API container

Full CI/CD planned

🧩 Use Cases

  • CRM sentiment monitoring

  • Social media monitoring tools

  • Customer feedback analysis

  • Product review analysis

  • Opinion mining for business analytics

📈 Impact & Outcomes

  • lexi.dev enables businesses and CRM teams to:

  • Identify customer emotions in real time

  • Detect negative feedback early

  • Improve customer retention

  • Enhance data-driven decision-making

  • Optimize support workflows using AI

👨‍💻 Author

Ashmit Dutta

🔗 GitHub Repository 👉 https://github.com/ftashmit/lexi.dev

📜 License

This project is licensed under the MIT License — feel free to use, modify, and distribute with attribution.

⭐ If you like this project

Don’t forget to star the repository and share it with others!