Test and adapt quickly: how Expedia develops for the cloud

As a global travel brand, Expedia operates localised sites in over 31 countries with over 40 million monthly visitors, where customers can be found logging in and checking details, checking in for flights or booking travel round the clock – and when you’re rushing to the airport or desperate to bag that bargain holiday – you simply don’t want to be kept waiting.

Being available or not can literally make or ruin someone’s long-anticipated holiday, so we need to be available 24/7 to address those needs. Back in the 90s we broke the mould in the travel market blowing the market wide open to a far wider range of choices for customers. What this means today is that we can process an infinite range of potential travel combinations every second.

[easy-tweet tweet=”Live streaming analytics and cloud-based algorithms are important to @expedia” user=”comparethecloud” usehashtags=”no”]

To ensure continuous development, incorporating 24-hour uptime and to support multiple language capabilities a global workforce spanning multiple time zones is a must, and developer teams must have the ability to act as one. This is why tools such as live streaming analytics and cloud-based algorithms are important to us: our need is for both near real time decision-making, and the cloud allows us to service our customers globally with fast scaling capacity. 

using the cloud to store queueing patterns and essential testing functionality becomes critically important

At Expedia developer teams work on everything from fast-moving travel requests to agile development, improving Expedia’s mobile App and mobile UX, to an on-the-job analytics experience to develop the “tech of tomorrow” or platform-related improvements touching most engineering teams’ core components, to algorithms that help our online marketing activities. With developers based everywhere from the USA to Australia, and everywhere in between, using the cloud to store queueing patterns and essential testing functionality becomes critically important.

Using the cloud, each team can work independently, whilst still collaborating as a larger team. Engineering and testing/hosting in the cloud is crucial to enabling 24/7 uptime and availability across all of our various systems. One team based in GMT can pick something up or escalate to another team as they go to sleep for an AST time zone based team to deliver on.

We’ve been in the travel technology business for nearly twenty years, but the last half decade has seen an unparalleled level of innovation. Global travellers are becoming savvier with each passing year and to continue stay ahead in the market and deliver on their evolving expectations, our teams are constantly tweaking and adjusting the data we make available to our travel customers in a way that is both reliable and provides what they need, whilst also optimising our business critical operations and stack architecture on the back end.

Back in the 90s the flexibility and choice revolutionised the travel industry

Back in the 90s the flexibility and choice revolutionised the travel industry. Today’s customers expect to navigate through these choices in seconds on any device and predictive and streaming analytics are crucial to helping customers to do. Cloud developers are increasingly becoming heroes in today’s travel technology workplace. Thanks to cloud hosting, big data capabilities as well as predictive and streaming analytics, developers across the Expedia ecosystem have the ability to do things today that they could not do only 5 or so years ago.

[easy-tweet tweet=”Cloud developers are increasingly becoming heroes in today’s travel technology workplace” user=”expedia” usehashtags=”no”]

In order to maintain our global market leadership a commitment to continual innovation is essential. Our developers’ main challenge now rests on the ability to test and learn faster than ever before – and that can mean failing fast in order to learn – whilst at the same time discovering the best possible consumer products they can build. Our developers are encouraged to try things at a rapid pace, to try, test, test again, and if/when the test is not successful to learn the essential lessons of what’s failed fast, and then move items along in a queue to the next task. 

We strive not to control or restrain our engineers – but rather to focus on fast recovery and information sharing with our global teams in diverse locations. It’s also vitally important that our teams can share best practice learnings across platform. So in addition to testing and learning with existing platforms, we encourage our teams to test new tools. The majority of our development stack teams work in Java, but our developers ultimately choose the languages that fit and also the manner in which they wish to deploy, then it’s a case of making sense of the scale of the development which needs to occur. Often we use a combination of tools for our Big Data and predictive analytics projects – whether it is Cloud/AWS, Java/Scala, PHP, Python or even Heroku or internal cloud infrastructure for others – we aim to develop an improved experience for what our customer sees when they reach our sites. 

[easy-tweet tweet=”We strive not to control or restrain our engineers says @Expedia” user=”comparethecloud” usehashtags=”no”]

Whether it’s a myriad of travel routes someone can take to get to a destination, or the number of popular queries a service receives, we try to bring all data into product development, and aim to repeat test features before we roll them out across the business. But it’s accomplished in a way that consistently keeps the goalposts clear: we aim to develop in a truly traveller-assisting way. 

Just as our customer’s feedback helps us to improve our services for future travel needs, we often live-test on site so that we don’t have to debate what customers want, but rather we can test live and build on data rather than speculation. The scale of the opportunity, and what we aim to achieve is great: often the traveller accessing our service does not even realise they are subject to a live test, or that they are helping to inform how often someone else can choose the London – New York route at 14:57 on a Tuesday. The fact is we aim to constantly, and consistently, improve on valuable travel experiences our customers have in a real-time way.

Given the importance of data and accessibility to that data, we imagine a hybrid approach of in-house and cloud platform services will remain one of the most effective ways we can do this successfully – particularly for an industry with as much decision-making and emotional needs as travel. So no matter if we are planning for executive travel needs or new millennials who are planning their own round-the-world trips, we aim to make our service delivery one that provides only the best traveller options available through current technology available to us.

+ posts

AI Show - Episode 1 - Clare Walsh

Newsletter

Related articles

Next-Generation Ransomware Protection for the Cloud

Have you heard the story of the LockBit gang?...

The art and science of GenAI in software development

The past year has shown just how quickly technologies...

From the lab to production: Driving GenAI business success

Generative AI (GenAI) was the technology story of 2023....

AI Show – Episode 1 – Clare Walsh

We invite you to join us on an extraordinary...

Generative AI and the copyright conundrum

In the last days of 2023, The New York...

Subscribe to our Newsletter