Background
Headquartered in Asia, a top global airline is a leader in the aviation industry and a member of the one of the largest international alliances. The airline’s executives committed to making Artificial Intelligence a business priority across the company and specifically were seeking an AI platform that leverages advancements in machine learning and deep learning to improve the airline’s ability to accurately forecast passenger demand for domestic and international routes.
The Challenge
With so much historical and real-time data available, the airline was looking to artificial intelligence to derive insights that could benefit various areas of the business including demand planning and customer service. Yet with volumes of data coming from a multitude of sources—the airline industry alone is predicted to generate 98 million terabytes of data by 2026—it was impossible to discover and apply these insights at scale with the legacy systems in place.
Profitability in the airline industry is rooted in an airline’s ability to accurately predict future demand and optimize routes accordingly. As every commercial flight operator knows, operating routes with little demand can result in major financial losses. The airline was looking to boost revenue by building and implementing dynamic pricing models for routes flown to hundreds of destinations.
Another major challenge presented itself: It wasn’t possible with its current team and legacy systems to quickly produce demand forecasting models with high accuracy rates. The process was labor intensive and required deep expertise in machine learning. It took months to build, test deploy and optimize models—time the company simply not afford to waste.
Adding to the complexity brought by managing volumes of data, the airlines relied on external IT firms to manage their infrastructure and business-critical software, which hindered their ability to collect the data in one place for model creation and optimization.
The Solution
After researching the potential of automated machine learning (AutoML) and deep learning (AutoDL) platforms, the leadership team identified OneClick.ai as the platform that could dramatically improve its ability to dynamically price routes based on demand forecasts.
Working with its outsourced IT team, the airline was able to generate a forecasting model that provided up to 90% accuracy for route demand planning in just a few hours. This enabled the airline to optimize pricing for routes based on the demand. This precise view of future passenger demand maximized profits for the company while saving time and resources.
Key Benefits
- Ease of use: Armed with historical and external data, the airline’s demand planning team was able to quickly build and implement a new AI forecasting model that outperformed the legacy model.
- Real-time Insights: The live view of how demand changes enabled the company to optimize route pricing in real-time.
- Major gains in accuracy: The 90% accuracy rate beat the accuracy of previous models by more than 10%, allowing for better planning that led to bottom-line gains.