Ant International’s FalconTST 2.0 adopted by global banks

Barclays, Citi, Deutsche Bank and Standard Chartered have integrated Ant International’s Falcon Time‑Series Transformer (TST) 2.0 into their systems to improve cashflow forecasting and foreign‑exchange liquidity management for cross‑border payments.

Ant International built FalconTST 2.0 for time‑series numerical data such as transaction amounts, account balances, settlement flows and currency positions. The company says the model learns cycles, trends, seasonality and sudden shifts by training across datasets from finance, retail, energy, travel and macroeconomic indicators.

Banks use the model to predict when funds will be needed, how much is required and in which currencies. Those forecasts feed intraday liquidity planning, capital allocation and FX exposure management in cross‑border payment operations.

Ant has positioned FalconTST to handle multiple forecasting tasks that previously required separate models. The company says the single model can support sales forecasting for retailers, demand planning for airlines and liquidity forecasting for financial institutions.

Ant plans additional applications for FalconTST, including demand forecasting for e‑commerce supply chains and predictive operations management in aviation. Banks that have integrated the model report using it for liquidity exposure forecasting and FX management.

“Large language models have shown how AI can understand and generate information,” commented Jiang‑Ming Yang, chief innovation officer at Ant International. He added that FalconTST focuses on how numerical flows change over time and on converting forecasts into operational decisions about liquidity, FX exposure and capital allocation.

Banks report the model helps reduce unexpected funding shortfalls, improve management of currency positions and align capital with real‑time payment flows.