Citi and Ant International have launched a pilot program using artificial intelligence to help airline customers cut foreign exchange (FX) hedging costs on online ticket sales, the companies said on Friday.
The pilot integrates Ant International’s Falcon Time-Series Transformer (TST) Model with Citi’s Fixed FX Rates solution. The AI model, developed by Ant International, uses deep learning and time series forecasting to improve cashflow and FX exposure predictions for businesses with multi-currency sales.
“This is the first industry-tailored solution developed from our Falcon TST Model with a bank partner,” Kelvin Li, General Manager of Platform Tech at Ant International, said in a statement. “The 30% hedging cost savings Ant International has achieved for the pilot airline customer shows the cost efficiency that can be achieved with AI-enabled FX hedging.”
One of Asia’s leading airlines has already deployed the solution, Citi said, and recorded reduced FX hedging costs during initial live transactions.
The combined solution addresses currency fluctuation risks by locking in FX rates across more than 70 currencies, offering pricing predictability for e-commerce transactions. Citi’s Fixed FX Rates product is already widely used in sectors such as airlines, travel, and online commerce.
“We are constantly evolving our products and solutions to support new use cases built on our clients’ priorities,” said Sam Hewson, Global Head of FX Sales at Citi. “This innovative solution sees us leveraging best-in-class technology capabilities in the broader ecosystem to accelerate go-to-market use cases.”
Ant International’s Falcon TST Model uses a transformer architecture with nearly 2 billion parameters. The company reports achieving forecast accuracy rates of over 90% in its own applications, a figure it expects to maintain as the solution expands across industries.
Citi and Ant International said they plan to extend the solution beyond airlines to serve other e-commerce and travel industry clients.
Business News Asia
