Tabular data is everywhere in the corporate world. I feel like there is a zillion opportunities to make money with these transformers. But how to enter that market remains a mystery to me.

> applications like customer analytics, fraud detection, and supply chain optimization

Part of a graph rewrite engine? Of an optimizing compiler? Complex Event Processing?

Why would I need to use this over a flexible multi-agent framework? What’s the benefit?

Exciting to see end-to-end deep learning go head-to-head with XGBoost and random forest.

Can someone explain the use case for this