Wells Fargo built an in-house foundation model using billions of tabular customer transactions rather than text. The system outperforms traditional machine learning models in fraud detection, reduces costly false declines, and now serves as a shared backbone for credit scoring and compliance.