TruvaLI AI
TruvaLI is a language model we trained ourselves for AML, fraud and compliance work. TruvaLI AI is where anyone can use it for free, and where compliance professionals help decide what it learns next.
Three names, one project
- TruvaLI is the model: a vision-language model Bitrelic trained for compliance work.
- TruvaLI AI is the free chat at chat.truvali.ai and the community of compliance professionals around it.
- TruvaLI AML is our platform for sanctions screening, transaction monitoring, identity verification and risk scoring.
The platform is a product institutions buy. The model is something we want the whole compliance profession to be able to use, question and improve. The pages below explain what TruvaLI is, how it was built and where it is going.
TruvaLI, our compliance model
TruvaLI is a language model trained for the daily work of compliance teams: regulation questions with citations, typologies and red flags, adverse media and …
Read →
TruvaLI technical report
This report describes how TruvaLI is built: the source archive, how training examples are generated and filtered, how the model is trained and served, …
Read →
Community and social responsibility
Financial crime is not only a risk on an institution's balance sheet. It has victims. We made TruvaLI free because every compliance officer who …
Read →