Traveloka was founded with the goal of presenting the best travel plans to each user of its popular mobile app. From the beginning, Traveloka has been dedicated to building a data-driven business. Their world-class data science team strives to develop accurate, flexible, and high-performing predictive models that create a compelling user experience for researching flights, hotels, and travel-related services.
But as the popularity of their mobile app grew, the Traveloka data science team were becoming victims of the company’s success. Initially, machine learning was a mostly manual process, with the team iterating on several models, trying to find the one that would result in the optimal travel plan for each customer. Results were very good, and the company was booming.
The team was frustrated with the time required to manually tune even a limited number of models, as it hampered their ability to develop a truly agile data science environment, and caused issues like:
Arditto was keenly aware that their current methodology was not the best way to help his fast-paced company grow.
DataRobot worked with Arditto and his fellow data scientists to determine how DataRobot could meet Traveloka’s needs. Early in the investigations, it became apparent that DataRobot was extremely effective in producing highly accurate models for Traveloka in very short order.
DataRobot enables fast development, testing, and iteration of dozens of open source models, and offers built-in evaluation tools to guarantee model performance. In addition, DataRobot delivers metrics that visually explain why a certain model is best. This, in turn, allows the data scientists to easily, and visually, communicate with management.
Deployment is often the most difficult step in a predictive modeling project. In Traveloka’s case, deployment requirements were forcing trade-offs between model performance and accuracy, or what they call “agility” – the ability and time required to deploy predictive models.
Arditto now uses the DataRobot Prediction API to deploy models into production at supersonic speed for truly agile machine learning development. This new-found agility reduces business stakeholders’ concerns about resources and delivers a positive impact to the business. It was, in fact, a major component in the decision to purchase.
|Traits & Capabilities|
|Traits & Capabilities||Data Scientist Mortal||Data Scientist Superhero|
|Bottom Line Business Impact|
|Bottom Line Business Impact||$100K+ per year||$M's+ / Year|
|Model Engine Tested|
|Model Engine Tested||No||200,000,000+ times|
|Standardization||When there's time||Every time|
|Multitasking||Single model||Multiple models|
|Package installation and dependency management|
|Package installation and dependency management||Manual||Not needed|
|Risk of coding errors|
|Risk of coding errors||Substantial||None|
|Popularity with business users|
|Popularity with business users||Somewhat||Very|
|Business user engagement|
|Business user engagement||Too busy||Always available|
|Helped with guardrails|
|Helped with guardrails||None||Yes|
|Skills Needed||Significant - PhD-level||PhD not required|
|Model deployment effort|
|Model deployment effort||Significant||Effortless|
|Model iteration||Manual & slow||Automated|
|Helped by other Data Scientists|
|Helped by other Data Scientists||No||85+|
|Algorithms and libraries|
|Algorithms and libraries||A comfortable few||Many|
|Number of models / month|
|Number of models / month||Very Few||Thousands|
|Model Transparency||None||X-Ray Vision|
|Explainable Models||'Black Box'||Automatic|
Now, Arditto and the Traveloka data science team are looking to supercharge their efforts. By producing even more accurate models, they will provide Traveloka app users an even more customized user experience. And by continually evaluating and tuning their models, and looking for additional sources of data to refine their predictions, Traveloka will continue to outpace the competition for years to come.
“DataRobot had significantly reduced time spent on creating and tuning predictive models. I’ve been able to become far more productive and get models into service more quickly than ever before – all of which helps Traveloka deliver the best travel deals to our customers.”