Skip to content

Multi-site web ecosystem for CTE Trailers and Selectrailers

Multisite Web Ecosystem

Three separate sites unified into a single multi-region, multilingual platform, with automatic synchronization of vehicle inventory.

Project Sheet

Period
November 2023–August 2026
Client
CTE Trailers Romania, CTE Trailers Bulgaria, Selectrailers
Field
automotive, commercial vehicles
Our role
architecture, development, infrastructure, and data integration
Status
technical development completed, in pre-production; public launch follows the client marketing team's schedule

The Challenge

The client managed three separate WordPress sites, with information and translations distributed across multiple systems. The vehicle inventory was handled separately in an internal application.

Publishing and updating content for each region required repetitive operations. Inventory synchronization and multi-channel sales administration needed an automated and verifiable workflow.

What We Did

We designed and developed a platform with centralized administration, serving all sites, regions, and languages from the same infrastructure, while maintaining each brand’s domain and identity.

Content administration. The platform includes 35 content collections, 23 types of page blocks, preview for editors, SEO and hreflang configuration, redirects, and distinct permissions for each role and region.

Inventory synchronization. An automated workflow validates sources, detects changes, and uniformly applies creations, updates, and deletions. It synchronizes images, documents, and attributes, allowing publishing rules for each vehicle and sales channel.

Delivery and operation. The project includes automated testing, security scanning, separate development and pre-production environments, monitoring, and infrastructure defined as code.

AI agent access. A Model Context Protocol server gives AI agents controlled access to CMS operations, including authentication, permissions, and action logging.

With What We Built

The platform uses Payload CMS 3, Next.js 16, React 19, TypeScript, and PostgreSQL 17. AI agent access is implemented via Model Context Protocol and OAuth 2.1 with PKCE.

Infrastructure includes Docker, GitLab CI/CD, Portainer, Traefik, Cloudflare, Azure Container Apps, Azure PostgreSQL Flexible Server, Key Vault, Bicep, S3-compatible storage, Sentry, Playwright, and Vitest.

Result

Unified Sites
3, from 3 separate systems
Managed languages
3
Content collections
35
Page block types
23
AI agent tools
204
Unit and E2E tests
349
Development duration
13 months
Current use
the MCP server is used by the client’s marketing team via ChatGPT