Nimrod Kor builds developer tools and code analysis software. As the co-founder and CTO of Baz, he focuses on moving AI coding agents past simple autocomplete and into reliable development feedback. He previously shaped Infrastructure as Code security as the top contributor to the open-source project Checkov.
Nimrod regularly speaks about the architectural limits of LLMs. He explains why standard models fail at code review, breaking down how memory and reflection can fix their lack of judgment. His work challenges the idea that prompt engineering is enough to build genuinely trustworthy AI agents.
