Introduction

Most popular customer-centric OTA wants to build Right price at Right Time to purchase travel entitlements (Air, Hotel, car) in a very engaging fashion by leveraging the power that comes from combining massive amounts of data and machine learning.

• Create parallel processing for pulling data from multiple GDS

• Build Hotel connectivity for different end points

• Data aggregation based on trends

• High throughput, low latency and caching

• Building high performant thread safe code

• Complex algorithms to observe pricing trends

• Auto scaling and optimizing utilization of infrastructure

• Delivered a consistent connectivity platform by converging several connectivity integrations (eg: Expedia, QuickConnect, Agoda, Sabre & Get-a-room).

• Built abstractions that facilitate common patterns found while searching, selling and booking travel segments

• Realized auto scalable platforms through hybrid solutions and innovative persistent mechanisms.

 Technological Framework

Why this setup?

Scala and Apache Spark provide distributed, high-throughput processing across multiple GDS sources simultaneously, while Redis caching and Kinesis streaming ensure low-latency, real-time pricing data delivery.

Why this setup?

AWS EKS with Kubernetes enables automatic, component-level scaling of connectivity workloads based on real-time demand — ensuring infrastructure costs scale precisely with usage.

Why this Setup?

A unified abstraction layer over multiple GDS and hotel connectivity integrations ensures consistent search, sell, and booking patterns regardless of the underlying supplier system.

Takeaway

Zapcom delivered a unified, auto-scalable travel connectivity platform converging GDS and hotel integrations, achieving 20% infrastructure cost reduction and 5x improvement in integration time for connectivity.