SafaArban Translator is a professional, human-in-the-loop translation web application designed for accuracy and efficiency. This platform allows users to upload various document formats, extract text while preserving the original layout, and perform manual, line-by-line translations in a user-friendly bilingual editor. The application supports right-to-left (RTL) languages and provides tools like translation memory and a glossary to ensure consistency and quality.
Our goal is to build a production-ready web portal that streamlines the professional translation workflow. By focusing on human translation and providing robust tools, we aim to deliver a secure, scalable, and intuitive platform for language professionals. This project is built with a modular design to allow for future integration of self-hosted AI models, without relying on external AI APIs in its core version.
git clone https://github.com/your-username/safaarban-translator.git
cd safaarban-translator
npm install
.env file in the backend directory.# In the root directory
npm run dev
This will start both the frontend and backend development servers.
The project is organized as a monorepo with two main packages:
frontend/: The React application that provides the user interface.backend/: The Node.js/Express application that handles the business logic and API.Each package has its own package.json and scripts. The root package.json contains scripts for running both applications concurrently.