3Victors Uses AWS to Deliver Near-Real Time Travel Trend AI Insights

Some of the world’s leading travel technology companies rely on 3Victors to help determine how to best engage travelers and what specific marketing content to offer them to maximize engagement while optimizing profit. 3Victors ingests data from the world’s largest reservations systems to provide data analytics solutions to sellers of travel across the globe. The growing startup uses a proprietary platform of big data capture and artificial intelligence–based algorithms to process billions of shopper requests for flights every day. The company’s platform delivers actionable real-time insights to travel marketers and revenue managers, who use the insights to analyze the competitive landscape and market dynamics of air-related data and offers from travel marketers.

To provide the best business value to travel customers, 3Victors needed to easily aggregate the fast-growing volume of data coming in daily from disparate global reservations systems. “We knew we were going to grow way beyond a petabyte of data,” says Rick Seaney, the chief executive officer of 3Victors. “Our company also needed to decide whether to renew our facility agreement for the on-premises collocation facility that hosted our platform. We realized it was the right time to move to the cloud, which could give us near-infinite scalability.”

3Victors addressed its challenges by implementing an Amazon Web Services (AWS) solution to capture and store data in a data lake. “AWS offered the best technology for us when it came to scalability and data analytics, so it was an easy choice,” says Seaney.

3Victors uses Amazon Route 53 to direct vendor content to AWS Elastic Beanstalk, logging application-specific statistics to an Amazon Redshift data warehouse via Amazon Kinesis Data Firehose. The solution transforms captured pricing data and places it on a 24-hour Amazon Kinesis streaming buffer balanced across more than 100 databases. Two concurrent pipelines fan out from this streaming buffer and feed Elastic Beanstalk clusters that read and deserialize the data in near-real time. Data is sent to smaller data lakes using Amazon EMR, which is cost-optimized using Amazon Elastic Compute Cloud (Amazon EC2) Spot Instance fleets. By taking advantage of Spot Instances, 3Victors gains access to unused Amazon EC2 capacity at up to 90 percent less than On-Demand prices. Finally, an analytics pipeline extends out to unlimited Amazon Kinesis stream buffers that support an extensible micro-batching framework. 3Victors customers use data visualization tools to access data via serverless Amazon API Gateway and AWS Lambda deployments. Using this solution, 3Victors has an extensible, near-real time predictive analytics pipeline that enables demand forecasting and deal classification.

Using AWS, 3Victors can ingest more than a billion worldwide travel searches and 230 billion priced itineraries every day from multiple data sources, including the world’s largest reservations systems. “We are growing at a rate of 20 terabytes of data every day, and we would have had to quadruple the size of our infrastructure to support it in an on-premises environment. There’s no way we could afford to do that,” says Seaney. “AWS enables us to ingest and analyze as much data as we can.”

According to Seaney, the typical turnaround time for business intelligence data to airlines is up to 30 hours. “The problem is, of the more than one billion searches we receive each day, 38 percent of the itineraries significantly change by the time we get the data ready for customers. By processing in near-real time on AWS, we can deliver fresher data and insights so decisions can be made intraday, completely transforming current batch processing pipelines into next-generation predictive and reactive analytics,” says Seaney.

Because 3Victors can process more data faster, it can give its customers better insights. “The more data we process, the better and more accurate the answers are to the questions our customers have,” Seaney says. “We can help them improve the likelihood of their customers booking flights. Using AWS, we are simplifying big data to help our customers increase conversion and return on ad spend while optimizing revenues.”

3Victors is also reducing operational costs by moving completely to AWS and running its big data environment on the AWS Cloud. “It would be very difficult for us to do what we’re doing without a secondary market for compute,” Seaney says. “Amazon EC2 Spot Instances give us a huge advantage. We’re saving tens of thousands every month by using this solution. We couldn’t exist as a company, processing 230 billion travel itineraries daily, without the scalability and affordability of AWS.”

To learn more, visit aws.amazon.com/big-data/datalakes-and-analytics or aws.amazon.com/blogs/industries/a-tech-ceos-journey-from-cloud-skeptic-to-evangelist.