Al-Qayser Group. Three businesses, independently run.
Al-Qayser Group needed working platforms behind three bilingual digital products: design and real estate, car wash, and dry cleaning. We built one AWS reference architecture and deployed it as three separate environments, each with its own data and operations.
Three prototypes needed a working backend.
The three products ran in the browser without the services needed to launch. Property enquiries had no durable record, car-wash bookings had no central capacity control, and dry-cleaning orders could not reach the store's desktop.
Each business needed secure hosting, a release process, monitoring and a private link to its premises. The group also wanted separate ownership and operations, so changes in one business would not affect the others.
A common foundation with separate responsibilities.
Separate the businesses at the account boundary
Each platform has its own AWS account, data and environments. Reusable infrastructure code keeps the foundations consistent while allowing independent billing, scaling and operations.
Give every request a system of record
PostgreSQL stores enquiries, bookings, memberships and orders. Transactional booking enforces the car wash's three-cars-per-slot limit, while queues carry notifications and operational updates.
Connect the premises privately
Site-to-Site VPN links the head office, site desk and dry-cleaning store to their respective platforms. Internal tools and client documents stay behind controlled access.
Make releases and recovery repeatable
ECS Fargate runs the application services, with a Multi-AZ database and encrypted media storage. Automated deployment, rollback tests, monitoring and runbooks support each business after handover.
From browser prototypes to operating platforms.
All three businesses received production platforms with central records, automated notifications and independent operations. Car-wash load tests checked booking integrity at five times expected peak demand, and handover included runbooks and an infrastructure workshop.
Availability remains a 99.9% design target. Recovery objectives were set at a recovery point of five minutes or less and recovery within one hour; they should not be read as a historical availability record.