Olaoluwa Shaun Faseru

I have spent my career managing and developing machine learning models into production where mistakes are expensive. In financial crime at a UK bank, my work is transaction monitoring: anomaly detection, risk scoring and the daily data pipelines that feed them, at the scale of millions of transactions.

Most of that work is about trust. A model that flags everything is ignored, and a model nobody can explain will not pass review. The craft is in calibrated scores, alerts sized to the team that reviews them, and evidence a regulator will accept.

AI agents now need the same treatment. They act with real credentials, and nobody is watching them the way banks watch payments. I founded BARA Labs to build that monitoring, and to build it in the open.