Enable next-generation retailing with a ONE Order system built on AWS

ONE Order, as defined by the International Air Transport Association (IATA), is an industry-led initiative intended to simplify airline reservation, delivery, and accounting systems. It works to gradually phase out the current booking, passenger name records (PNRs), ticketing records, etickets, and electronic miscellaneous documents (EMDs). As part of this directive, IATA has documented the messaging standards, processes, and implementation guidelines for the ONE Order system.

This directive is a major step toward next-generation airline retailing, emulating the retail industry very closely. This frees airlines from complex processes and simplifies data storage, sharing, and interaction with partners in the retailing ecosystem, including revenue settlement.

In my current role, I have had the opportunity to have extensive conversations with airlines, airports, and hotel customers. During these conversations, airlines have shared that they are looking to build or buy a ONE Order–compliant order-management system and integrate it with their existing ecosystem. In this blog, I want to share ideas for building a ONE Order system natively on Amazon Web Services (AWS) and integrating that system with an airline’s IT and partner ecosystem.

ONE Order defined

The ONE Order directive simplifies the processes of order creation, management, and tracking, irrespective of different fulfillment and marketing partners. It makes the payment-settlement processing between parties much quicker and helps parties to view the changes to orders in near real time across the ecosystem, improving operational efficiency and the customer experience.

As part of this directive, IATA has defined the following:

• XML standards to model the order data structure, which contains all the details of the customer order, closely resembling a typical order in the retail industry
  • This replaces the PNR, EMD, and ETK data structures that exist currently in most airline systems.
• business events that model the lifecycle of the order and push order lifecycle events to other systems of interest
  • This could be internal IT systems—revenue-management and fulfillment systems—or external partners such as code-share partners and nonair product/service providers like hotels, land products, and retail.
• implementation guidance on how the new set of processes span between parties in the order ecosystem

Building a ONE Order system natively on AWS

The following diagram captures a potential architecture for building a ONE Order system natively on AWS.

The above design focuses on creating an event-driven architecture that facilitates a seamless information exchange between the channels, the order store, other IT systems, and the partners in the ecosystem. Provisions are made to monitor order key performance indicators (KPIs) in near real time and publish the information to an enterprise data lake for further insights. Coexistence between ONE Order systems and the traditional PNR-based passenger service systems (PSS) is also addressed.

The core order data store is built using Amazon DynamoDB, which is a fast, flexible NoSQL database service for single-digit millisecond performance at virtually any scale. Amazon DynamoDB supports encryption at rest, automatic backup and restore, and guaranteed reliability with a service-level agreement (SLA) of up to 99.999 percent availability. DynamoDB Streams captures a time-ordered sequence of item-level modifications in the order table and streams this information for other applications and partners to know about orders being created, modified, fulfilled, or cancelled. Order services are deployed as container workloads on Amazon Elastic Container Service (Amazon ECS), a fully managed container orchestration service, and API end points are securely exposed through the Amazon API Gateway, a fully managed service that makes it easy for developers to create, publish, maintain, monitor, and secure APIs at virtually any scale.

The core messaging between the various components is realized using Amazon Managed Streaming for Apache Kafka (Amazon MSK). Using Amazon MSK, customers can greatly reduce operational overhead, including the provision, configuration, and maintenance of highly available Apache Kafka and Kafka Connect clusters.

Orders and aggregates are stored in Amazon OpenSearch Service—which makes it easy to perform interactive log analytics, near-real-time application monitoring, website search, and more—for near-real-time, read-only access and dashboarding through Amazon Managed Grafana, which offers scalable and secure data visualization.

Individual transformation logic for transformation between formats or initiating business events based on the order lifecycle events are built using AWS Lambda, which is a serverless compute service for running code without having to provision or manage servers. You pay only for the compute time that you consume.

This architecture allows for coexistence with the existing PNR-based PSS systems by providing a two-way event exchange mechanism that ensures synchronization of the order data with any changes that happen in the old PSS system.

Integration with airline IT and partner ecosystems

The primary integration of the ONE Order system with the airline IT and partner ecosystems is either through an event or an API. APIs help other systems directly interact with ONE Order systems to create, modify, and cancel orders. Typical airline customer channels (direct and indirect) use this mechanism to create, modify, or cancel orders.

Order lifecycle events defined by IATA can enable other systems to be notified in case any new orders are created, changed, or canceled. These events help other airline IT systems, including revenue-management, PNR-based PSS systems, and inventory-management systems, understand what has happened to the order, evaluate implications, and kick-start downstream processing.

These events also help integrate with other partners in the ecosystem, such as other service providers and fulfillment agents like hotels, airport services, and other nonair retail providers that are jointly sold by the airline.

Usage of events also ensures a decoupled, asynchronous processing, making this architecture scalable and highly resilient.

Next-generation retailing

For those who have worked in or with airlines for several years, like I have, it’s exciting to be a part of the evolution of the industry’s retailing and personalization efforts. I am looking forward to working with airlines that want to build next-generation retailing platforms natively on AWS that adhere to IATA ONE Order standards.

As a continuation of this blog, I have detailed the integration between the ONE Order system and revenue-management systems in airlines, which you can read about in the AWS for RMS: Modern revenue management in the cloud ebook.