Laith Rastanawi
Software Consultant & Mathematician
Berlin-based software consultant and mathematician, working on software systems and AI integration for enterprise clients. PhD in Mathematics from Freie Universität Berlin, with research in computational geometry and discrete mathematics. Transitioning into AI safety, with a focus on evaluation, failure modes, and monitoring of advanced AI systems.
Updates
-
Completed BlueDot Impact's Technical AI Safety Project Sprint, building Come Clean, an evaluation framework testing whether coding agents disclose accessing restricted information in their self-reports. · Certificate
-
Accepted into the Fall 2026 round of SPAR (Supervised Program for Alignment Research), selected for "Invisible Hands: Measuring Agent Steering of Human Researchers," mentored by Trevor Lohrbeer.
-
Attended the 2026 European Frontier AI Safety Day in Berlin, gathering Europe's AI safety community for keynotes, workshops, and panel discussions.
-
Gave a talk, "Open Foundation: Open Source and Open Standards for Digital Infrastructure," at SAFIERR-TECH 2026 in Latakia, Syria. · LinkedIn Post
© SAFIERR TECH 
© SAFIERR TECH -
Published AgentAudit, a static security analysis tool for AI coding agent sessions. It analyzes tool calls against configurable rule-based patterns to identify suspicious activity. · Demo
