Travello App: Delivering a “market-of-one” mobile application for travelers on Google Cloud

About Travello App

Travello App is a social application that enables travelers to connect with like-minded people in destinations they are visiting.

Industries: Travel & Hospitality
Location: Australia

Using Google Cloud Platform, Travello App delivered a highly personalized app that enables travelers to contact others heading to or at the same destinations and view services such as local tours, activities, and hotels.

Google Cloud Results

  • Runs a highly personalized mobile application used in 180 countries
  • Supported growth within two years to over two million feed events and tens of thousands of new connections
  • Accommodates traffic peaks without any deterioration in speeds

Achieves application response times as fast as half a second

Meeting new people is a welcoming part of the experience for traveling Australians. However, according to Ryan Hanly, there was no “really good” way to connect digitally with other people while on the road. So Hanly and fellow avid traveler Mark Cantoni decided to create a mobile application that would operate as a social network for travelers.

“We started planning an application that would be highly personalized—in other words, tailored to a market of one,” say Ryan Hanly, CEO and Co-founder, Travello App. “We wanted to curate the application around travelers, their interests, and locations.”

Hanly and Cantoni wanted the application to enable travelers to enter details themselves or authorize the collection of their information from social channels. The application would then use that information to connect users with like-minded travelers.

“Our feed algorithm would be based around your interests and locations, so we would show you posts from places or locations you were going to,” Hanly says. “We also wanted to add features that would, for example, enable you to add planned trips to your profile and to see who is expected to head to the same destination at the same time.”

Scalability and a global presence key priorities

The business partners established scalability and a global presence as key priorities for the platform that would run the application. Following experimentation and development in a Parse environment, the business decided to deploy the application—called Travello App—in a microservices architecture. This architecture would run in a cloud environment that supported open source technologies.

“Google Cloud Platform—and Google Kubernetes Engine in particular—was designed perfectly for our mobile app and web presence business as it had the flexibility and scalability we needed."

Richard Barge, Technical Lead, Travello App

After reviewing available cloud services, the founders decided to deliver the application on Google Cloud Platform (GCP). “What really brought us to Google Cloud Platform was Google Kubernetes Engine, the Kubernetes-powered managed environment for delivering Kubernetes applications,” Richard Barge, Technical Lead, Travello App, says. “Google Cloud Platform—and Google Kubernetes Engine in particular—was designed perfectly for our mobile app and web presence business as it had the flexibility and scalability we needed.”

GCP also met the Travello App requirements for a low-cost infrastructure. “Google Cloud Platform was a significantly better value than alternative options and its range of services was a closer fit for us,” Barge says. “In addition, we didn’t want to be locked into any one cloud services provider. Google allowed us to use the resources and products we wanted when we needed them.”

Travello App then deployed its application on GCP. “The move took a few months, but most of this time was dedicated to rewriting code rather than actually shifting to Google Cloud Platform,” Barge says.

“Because we were moving to Kubernetes and Google Kubernetes Engine, there was a lot of help freely available on the internet and the process itself was quite intuitive.”

Richard Barge, Technical Lead, Travello App

Intuitive process aided migration

The Travello App team completed the project using internal resources and expertise. “Because we were moving to Kubernetes and Google Kubernetes Engine, there was a lot of help freely available on the internet and the process itself was quite intuitive,” Barge adds.

After initially running the application on Google Compute Engine and Google Cloud SQL, Travello App completed the move of the application to a microservices architecture. Google Cloud Pub/Sub message-oriented middleware links each microservice on the platform.

The business now uses Google Cloud Natural Language API to extract information and sentiment about locations from travelers’ social media feeds, while Google Cloud Vision API classifies images for travelers to view relevant photos of cities, beaches, or other attractions.

Google Cloud Vision API also enables Travello App to filter out any inappropriate content, while Google Stackdriver provides monitoring and logging capabilities to help ensure the Travello App applications and architecture are running in line with business requirements.

“Thanks to Google Kubernetes Engine, our deployment process is now seamless and foolproof, with no user downtime.”

Richard Barge, Technical Lead, Travello App

Highly automated delivery

“Because our architecture is Kubernetes-based and supports a continuous delivery system, any code committed to the deployment process is automatically pushed into staging,” Barge says. “Pushing the code from staging to production simply involves updating Docker version numbers. So thanks to Google Kubernetes Engine, our deployment process is now seamless and foolproof, with no user downtime.”

Using Google Kubernetes Engine has also enabled Travello App to maintain a small team without DevOps or system administration specialists. This team can focus almost solely on application development and associated activities aimed at expanding the business and improving the user experience.

Rapid response times

“Our key focus is on response times for our users,” Barge explains. “Because feed data is curated for each person, our back-end systems generate a lot of activity in matching feed posts to travelers’ interests. We’ve been aiming to keep response times between half a second and one second and—based on data obtained through the Google Stackdriver Trace API—we’re achieving that. We’re quite proud of that outcome.”

Meanwhile, autoscaling delivered through GCP accommodates traffic peaks generated during marketing campaigns without any deterioration in response times. While the Travello App GCP environment is located in the us-central1 region, Google Cloud Load Balancing helps ensure response times in all markets remain fast and well within business requirements.

Supporting more than 2 million feed events

The GCP architecture for Travello App now supports over 2 million feed events and tens of thousands of new connections. Looking forward, the business is excited by the potential of Google Cloud Machine Learning Engine and artificial intelligence to deliver more predictive capabilities for the application.

“We want to learn what people want to see almost before they do so themselves,” Barge says. “That will be particularly important as we extend the range of services we’re showing customers beyond tours, activities, and hotels.”

Travello App is now preparing to build on its early high-growth phase. “We’re still in the early stages of where we want to be,” Hanly says. “We’ve got a global domination focus and there’s huge room for growth in our niche industry vertical. While we’re used in 180 countries already, we aim to increase our density of use within key markets such as the United States and Asia,” he adds.

About Travello App

Travello App is a social application that enables travelers to connect with like-minded people in destinations they are visiting.

Industries: Travel & Hospitality
Location: Australia
Google Cloud Platform logo

12 Months FREE TRIAL

Try Kubernetes Engine, BigQuery, and other Cloud Platform products with $300 in free credit and 12 months.

TRY IT FREE
Google Cloud Platform logo

12 Months FREE TRIAL

Try Kubernetes Engine, BigQuery, and other Cloud Platform products with $300 in free credit and 12 months.

TRY IT FREE