Summary
Sultan Hassan is an AI safety researcher and senior staff scientist with over a decade of experience building ML systems that accelerate scientific discovery, particularly across astrophysics and high-dimensional simulation domains. He combines hands-on engineering—leading Agile DevOps teams and shipping LLMs and agentic tools at the Space Telescope Science Institute—with deep research in causal inference, OOD generalization, and robust uncertainty quantification. Sultan has repeatedly translated ideas into impactful tools (e.g., CosmoGemma, an AI research assistant with >95% accuracy and thousands of visits) and achieved dramatic speedups in simulation and generative modeling pipelines. A former NASA Hubble and Flatiron/Simons fellow, he pairs fellowship-funded independent research with mentorship of students to productionize emulators and diffusion models. Based in Baltimore, he is currently focused on improving model reliability through localized fine-tuning and probe-based LLM monitoring in collaborative AI-safety projects.
10 years of coding experience
Master's degree Astrophysics, Master's degree Astrophysics at University of Cape Town
Bachelor's degree Physics, Bachelor's degree Physics at University of Khartoum
Doctor of Philosophy - PhD Astrophysics, Doctor of Philosophy - PhD Astrophysics at University of the Western Cape