AI Research Lead
Lukas Brandt
Berlin, Germany
About
Lukas leads the research team responsible for Arty's agent reasoning layer and the evaluation framework used to qualify new agents before they enter the library. He holds a PhD in machine learning from TU Berlin and previously worked on applied NLP at a European fintech, where his work reduced false positive rates in fraud detection by 34%.
"Most AI failures in production aren't model failures — they're architecture failures. The research question that matters is: how do we make agents that know when to stop and ask?"
Other team members
