Matthias Niehoff heads up data and AI architecture at codecentric AG. He helps teams design systems that actually work in production, shifting the focus away from standalone machine learning models to the less flashy but crucial infrastructure and organization that make data projects succeed.
He regularly speaks on how modern data workflows must adopt dependable software engineering practices. His talks break down complex topics—from demystifying storage formats like Parquet and Iceberg to arguing for simplicity in an ecosystem crowded with MLOps tools and GenAI hype. Matthias advocates for building what is necessary, rather than just what is new.
