AWS Case Study: Olery
Reputation is vitally important in the leisure and hospitality industry. Olery, an Amsterdam-based start-up, provides reputation management software for hotels that can turn customer feedback, online reviews and social media feedback into actionable business intelligence. Founded in 2010, Olery’s customers include leading hotels such as Boscolo Hotels, Okura Amsterdam, Van der Valk Hotels, Carlton Hotels, Best Western and Hampshire Hotels.
Olery’s products continuously monitor, gather, and analyze data from customers, hotels, and social media channels, which requires considerable computing resources. “Before moving to the AWS Cloud, we were running multiple application components on a limited number of virtual machines and the components were disrupting each other,” says Wilco van Duinkerken, CTO of Olery. “For example, when our system was analyzing German reviews, throughput for the English reviews decreased. Resources to support historical data for a large hotel group affected visitors’ access to the Olery website.” The Olery team decided to look for a new solution.
Why Amazon Web Services
“Amazon Web Services (AWS) provides a cloud platform,” says van Duinkerken. “We found that other cloud service providers merely provided virtual machines. AWS gives us computing power as well as a platform of related tools and techniques that help us leverage the benefits of cloud computing.”
“We took advantage of the AWS Cloud to divide our system into many small components that communicate with each other,” van Duinkerken continues. “The availability of insightful metrics with Amazon CloudWatch and features like Auto Scaling based on Amazon SQS queues are great examples of how AWS provides tremendous extra value, as opposed to just buying virtual machines from other cloud providers.”
Olery uses Amazon Simple Queuing Service (Amazon SQS) with Amazon Simple Notification Service (Amazon SNS), Amazon Simple Storage Service (Amazon S3) and Amazon Relational Database Service (Amazon RDS) to create and manage its environment. The company typically runs about 20 Amazon Elastic Cloud Compute (Amazon EC2) instances on average and can scale to almost 1,500 Amazon EC2 instances for large-scale data analysis, compute-intensive language analysis, or reporting. Olery uses Auto Scaling to scale its Amazon EC2 capacity automatically just by filling the queues, creating a predictable analytics pipeline that accelerates throughput speed on demand.
“By using AWS, we can add large amounts of customers and hotel groups at the same time for analysis,” says van Duinkerken. “Moreover, we can add large data sources to the analysis very quickly so the data is always up-to-date. Adding new data sources means that we have to gather and analyze the historic hotel review data for over 500,000 hotels and millions of reviews. When time is of the essence, we have found that we can now do this in less than 24 hours.”
When Olery decided to move their environment to the AWS Cloud, it also gave the company the opportunity to review, automate and document the deployment and continuous integration of all of its system components. “Now we have a fully scalable platform with flexibility and the ability to scale,” van Duinkerken comments. “Up until a few years ago, this was only available to big corporations with deep pockets. By using AWS, we have found that we can combine rapid development with refreshing and new ideas without being held back by infrastructure limitations.”
To learn more about how AWS can help your big data needs, visit our Big Data details page: http://aws.amazon.com/big-data/.