Hi there,
In 1982, an accounting professor named William McCarthy had a radical idea: don't record accounting entries — record the economic events that generate them. That idea became the REA framework, and today it underlies FIBO, banking knowledge graphs, and every AI agent that needs to reason about what a financial transaction means.
📚 New Tutorial
REA Framework & Bank Ontology: A Complete Tutorial
- What REA (Resources, Events, Agents) is — and why it matters more in 2026 than it did in 1982
- The three primitives mapped to real banking: loans, deposits, wire transfers, securities trades
- Four REA relationships: duality, participation, stockflow, commitment
- Step-by-step: build a bank REA ontology in Turtle (.ttl) syntax
- Python + RDFLib walkthrough with full SPARQL queries
- Real-world loan lifecycle modelled end-to-end in REA
- FIBO alignment and how REA maps to the industry standard
- REA as a foundation for LLM agents and ML fraud detection in banking
- Five hands-on exercises to test your understanding
📰 Recent Posts
Bank AI Agents Have No Kill Switch — and the Data Proves It — Wolters Kluwer's H1 2026 survey of 230 U.S. bankers found 72% lack AI kill switches or failure reporting — a production crisis hiding in the exact workflows where agentic AI is being deployed fastest.
Why Fraud Rings Survive XGBoost — and How GNNs Stop Them — Row-based ML misses coordinated fraud rings — GNNs expose them by propagating relational signals through transaction graphs.
More posts dropping every day. Stay curious.
— Bhanu @ superml.dev