Mohamed builds autonomous agents to automate complex financial processes. As a Senior Software and AI Engineer at Mika in Berlin, he focuses on replacing rigid German tax bureaucracy with agentic workflows.
Before tackling accounting, Mohamed worked as an AI Research Scientist at TU Berlin, where he built machine learning models for pathology. His current work dissects the heavy lifting of multi-agent architectures, targeting high-stakes domains where even a 0.1% error rate is unacceptable. He writes about machine learning on his blog, The AI Episteme.