Traduora

Ever Traduora is a full‑stack, self‑hosted translation management system (TMS) built to give developers granular control over every stage of the localization pipeline. At its core, Traduora exposes a RESTful API that allows projects to be created, keys to be imported/exported in multiple formats (JSON, YAML, CSV, XLIFF 1.2, Gettext PO, Android XML, etc.), and translations to be updated on‑the‑fly. The web interface is a single‑page React application that communicates with a NestJS backend, enabling real‑time collaboration through WebSocket channels for instant notification of translation changes.

Architecture

• Frontend – React 17+ with TypeScript, leveraging Redux Toolkit for state management and Ant Design for UI components. The SPA communicates via HTTP/HTTPS to the backend and uses Socket.IO for live updates.
• Backend – NestJS (Node.js) written in TypeScript, following a modular architecture that separates domain logic (projects, keys, translations) into independent modules. Controllers expose a comprehensive REST API, while services encapsulate business rules and data access.
• Database – PostgreSQL 13+ serves as the primary datastore, with a normalized schema that supports multi‑tenant projects and fine‑grained permission models. Prisma ORM is used for type‑safe queries, migrations, and schema generation.
• Search – ElasticSearch or PostgreSQL full‑text search can be enabled to power the instant search feature across keys and translations.
• Background Jobs – BullMQ (Redis‑based) processes import/export jobs, translation approvals, and webhook notifications asynchronously.
• Containerization – A Docker Compose stack is shipped with the repository, exposing services such as traduora-api, frontend, postgres, and optional redis. The image is also available on Docker Hub (traduora/traduora) and can be deployed to Kubernetes via Helm charts.

Core Capabilities

• Import/Export: Bulk operations support flat and nested JSON/YAML, CSV, XLIFF 1.2, PO, Android XML, and more. Importers parse metadata (e.g., context tags) to preserve structure.
• API Automation: Endpoints for project CRUD, key management (/keys), translation updates (/translations/:id), and batch operations. Webhooks can be configured to notify CI/CD pipelines of translation changes.
• Real‑time Collaboration: Socket.IO events (translation.updated, project.invited) keep all team members in sync without polling.
• Permissions & Roles: Fine‑grained RBAC (Admin, Editor, Viewer) with per‑project and per‑key scopes.
• CLI Integration: While not official, the community provides a CLI (traduora-cli) that wraps the API for command‑line workflows.

Deployment & Infrastructure

Traduora is designed to run on any environment that supports Docker. The official docker-compose.yml includes all dependencies, making a 5‑minute setup possible. For production, the stack can be scaled horizontally: multiple API instances behind a load balancer, PostgreSQL in HA mode (Patroni or Patroni‑Plus), and Redis for job queues. Kubernetes manifests are available, enabling rolling updates, autoscaling, and integration with external secrets managers.

Integration & Extensibility

• Webhooks: Trigger on key or translation events, allowing integration with GitHub Actions, CI pipelines, or custom notification services.
• Plugins: The backend exposes a plugin interface where developers can hook into lifecycle events (onKeyImported, onTranslationApproved) to implement custom logic (e.g., profanity filtering or ML translation calls).
• SDKs: A TypeScript SDK is bundled with the repository, simplifying API consumption in client applications.
• Open APIs: Full OpenAPI (Swagger) documentation is auto‑generated, enabling third‑party developers to generate client libraries in any language.

Developer Experience

Documentation is hosted at docs.traduora.co and follows a typical developer‑first structure: quickstart, API reference, architecture guide, and contribution guidelines. The project is licensed under AGPL‑v3, encouraging community contributions while protecting the open‑source nature of the platform. The active GitHub repository receives frequent commits, and community discussions are available on issues and pull requests.

Use Cases

• Continuous Localization Pipelines: Integrate Traduora into a GitOps workflow where pull requests automatically trigger import jobs, translators edit via the web UI, and approved translations are pushed back to the repository.
• Enterprise Localization: Deploy a self‑hosted instance with fine‑grained permissions and audit logs for compliance‑driven organizations.
• Multilingual SaaS Products: Host a single Traduora instance that manages translations for multiple microservices, each with its own project and key set.

Advantages

Traduora offers a blend of performance, flexibility, and licensing that differentiates it from commercial TMS solutions. Its self‑hosted nature eliminates vendor lock‑in, while the AGPL‑v3 license ensures that any derivative work remains open. The stack’s modern tech choices (React, NestJS, PostgreSQL) provide a familiar development environment for full‑stack engineers. Scalability is baked in through Docker/Kubernetes support and Redis‑backed job queues, making it suitable for both small teams and large