IZBooking · CASE STUDY
Engineering a Real-Time Travel Commerce and Auction Platform on AWS
IZBooking
CUSTOMERCHALLENGE
IZBooking combines digital travel booking with auction-based pricing, limited-slot inventory and real-time bidding. This creates requirements beyond a conventional online travel platform: merchant-aware routing, concurrent bid processing, durable bid state, transactional consistency and timely customer notifications all need to work reliably under production-like load. The POC was designed to validate whether the proposed AWS architecture could support these workloads before a broader production rollout
SOLUTION
TitanBases designed a multi-tenant, event-driven AWS architecture that separates edge routing, application services, transactional data, real-time auction state and asynchronous workflows. Amazon CloudFront, AWS WAF and CloudFront Functions handle protected merchant-aware routing. Containerized application services support booking, auction, pricing, loyalty and review workloads. Aurora PostgreSQL provides transactional data storage, while Amazon MemoryDB for Redis supports low-latency auction state. EventBridge, SQS and Lambda decouple booking, auction and notification flows, with OpenSearch supporting search and review-related workloads.
VALIDATION FOCUS
- Validated a multi-tenant routing model for merchant-specific travel storefronts using edge-based request resolution.
- Designed the auction workflow to support ordered concurrent bidding, durable bid state and low-latency processing requirements.
- Defined event-driven booking, loyalty and notification flows to decouple customer-facing transactions from asynchronous processing.
- Established a production-readiness validation path covering database failover, infrastructure automation and representative workload testing.
TECHNOLOGIES
Amazon CloudFront · AWS WAF · Amazon ECS · Amazon Aurora PostgreSQL · Amazon MemoryDB for Redis · Amazon DynamoDB · Amazon OpenSearch Service · Amazon SQS · Amazon EventBridge · AWS Lambda · Terraform
