Considered the “Google Now” for travel, You Are My Guide (YAMGU) is an innovative web and mobile app that leverages big data and machine learning to help users plan, personalize, and optimize trips of all types and sizes. Travelers enter their destination, when, and for how long, and YAMGU provides an optimized daily itinerary. It also allows users to discover trendy, must-see spots and events in real-time by looking through photos and comments from locals.

The Challenge

YAMGU processes data from more than 5,000 websites and blogs, including TripAdvisor and social media platforms like Facebook and Twitter, to provide its more than 2,000 daily users with highly personalized, real-time travel experiences.

YAMGU’s relational database, Microsoft SQL Server, couldn’t handle the increasingly large volumes of dynamic structured and unstructured data pouring in from YAMGU’s rapidly growing user base. It also had latency issues – YAMGU had a flood of B2B user requests coming in for which YAMGU needed to generate quick reports, and the database couldn’t keep up.

To handle its growth, YAMGU needed a data management platform that could handle multi-data center active everywhere deployment on the cloud while remaining always on even in the face of node outages.

“We needed a technology that could scale quickly and easily to handle our fast-growing user base and data sources, and would be easy to manage because we didn’t have too many resources,” said Mauro Bennici, co-founder and CTO at YAMGU. Having enterprise-grade text search and enterprise-grade analytics to be able to find and analyze data quickly was also critical for YAMGU.

The Solution

The team at YAMGU evaluated various data management alternatives, including Hadoop, MongoDB, and ElasticSearch, and concluded that none of them could handle the large volume of real-time data coming from new and varying sources, let alone provide 100% uptime and real-time, enterprise-grade search and analytics.

In DataStax Enterprise, YAMGU found a data management platform that was not only extremely flexible and easily scalable but also always on with Apache Cassandra while still being able to provide key enterprise capabilities such as search, analytics, operations management, developer tooling, round-the-clock support, enterprise security, and other benefits.

YAMGU currently runs Cassandra on Microsoft Azure with 45 nodes split between 2 data centers that handle a combined total of 10 million transactions per day and 24 terabytes of transactional data.

The Results

Cassandra has put YAMGU on a solid path to growth with customers experiencing faster response times and 100% uptime. Even with a constantly growing number of users and data sources, Cassandra gives YAMGU the ability to deliver a powerful solution that can store massive volumes of data consistently without the concern of downtime and performance degradation.

In fact, with its success in Europe, the team at YAMGU is now taking the app to the American market and planning to add a dedicated data center to support this expansion. Cassandra’s multi-datacenter replication capability gives YAMGU the flexibility to grow without concern.

“We can create a Cassandra cluster in any region of the world and move it from the development stage to production environment within minutes,” says Bennici. “It’s that easy.”

With DSE Search, YAMGU has been able to deliver rich, location-based user experiences based on proximity search. With DSE Analytics, YAMGU has reduced the time needed to analyze data and deliver reports requested by its B2B users from 3 days to 22 minutes.

“We are able to search and analyze vast amounts of data in the same environment in real-time, without needing to move it to a separate system,” said Bennici. “This eliminates the risk of losing data in transit and allows us to serve our users with the information they need and exactly when they need it.”

Having a single, integrated data management platform that can deliver all of these capabilities (scalability, 100% uptime, search, analytics) has allowed YAMGU to focus on growing its powerful and innovative application instead of worrying about keeping nodes up and running or whom to contact for support. Think of all that extra time for innovating, building, and providing for its user base. That’s the power of a great data management system.