At CFTE’s UK–Singapore Exchange, Agnes Bastaert demonstrated what AI looks like when deployed at scale.
The Problem: FX Risk at Scale
Ant International processes:
- $1.4 trillion in cross-border flows
Each transaction carries:
- FX risk
- liquidity challenges
For many businesses, especially SMEs:
- FX volatility directly impacts viability
The Solution: Falcon TST
Falcon TST (Time Series Transformer):
- analyses real-world signals
- predicts transaction flows
- optimises liquidity and hedging
Inputs include:
- transaction data
- e-commerce activity
- travel patterns
From Prediction to Impact
Results:
- 93% forecast accuracy
- <1% error on critical cases
- 40% reduction in hedging costs (pilot)
The Key Insight: Collaboration
Rather than replacing banks:
- Ant provides intelligence
- banks execute transactions
This creates:
- stronger partnerships
- shared value
The Takeaway
AI at scale is not just about models.
It is about:
- integrating real-world data
- embedding intelligence into infrastructure
- collaborating across the ecosystem
