Afiya Ayman

I am a Data Scientist working on the Deep Learning & AI team at Shell, where I focus on applied machine learning and generative AI systems for scientific and engineering applications. My current work involves developing LLM-based pipelines for document understanding and information extraction, and translating domain knowledge into scalable, production-ready solutions in collaboration with domain scientists.

My research specializes in Automated Multi-Task Machine Learning (AutoMTL), data science, and AI for social good, focusing on scalable deep learning and statistical models that improve training efficiency and real-world impact. Over the past seven years, I have contributed to NSF- and DOE-funded research projects, applying machine learning to improve public transit operations and enhance energy efficiency. My PhD thesis investigates factors influencing MTL performance and introduces cost-effective affinity prediction strategies to optimize task grouping across domains including computer vision, tabular data, time series, and transportation. I am especially interested in bridging research and practice, building AI systems that are both principled and impactful in real-world settings.

Interests

• Automated Machine Learning (AutoML)
• Multi-Task Learning (MTL)
• AI for Social Good
• Data Science