RBC Borealis Details ATOM, Its Large Transaction Model Wired to LLMs via Agentic Workflows

VectorInst · x · 2026-09-26

At the Vector AI Summit, Golnoosh Samei, Senior Director of Research at RBC Borealis, detailed ATOM (Asynchronous Temporal Model), the bank's Large Transaction Model (LTM) built for real-time, high-stakes decisions over massive transaction data in financial services.

RBC connects ATOM to LLMs through agentic workflows, moving beyond basic chatbots to deliver personalized, secure enterprise insights — a representative example of the domain-specific model + LLM agent approach in banking.

Original post →

More from Companies & People

Companies & People channel →