Abdalla Mohamed
PhD Student at The University of Texas at Arlington
About
I am a PhD student in Computer Science at the University of Texas at Arlington, working in the area of relational deep learning and relational foundation models. My research focuses on relational deep learning and relational foundation models, with an emphasis on understanding how these models interact with real-world relational data and database workloads.
Education
Research
I am investigating inference-time sampling strategies for relational foundation models, with a focus on semantics-aware approaches that guide sampling toward more contextually relevant outputs. By incorporating relational semantics to inform the sampling process, the goal is to make more effective use of limited token budgets while improving the accuracy of model predictions in downstream relational learning tasks.
Work Experience
- Researching semantics-aware inference sampling for relational foundation models to improve prediction quality under constrained token budgets.
- Leveraging LLM-extracted relational semantics to guide inference-time sampling toward more structure-aware outputs.
- Targeting improved performance at equivalent sequence lengths and baseline matching at reduced token cost.
- Served as Teaching Assistant for Database Systems and File Structures.
- Designed and delivered hands-on CS and robotics workshops for audiences from elementary students to working professionals.
- Developed curricula and technical materials covering programming, algorithms, and robotics systems.
- Diagnosed and fixed robotics hardware/software issues (sensors, control, integration) to ensure reliable lab sessions.
- Built and maintained Django-based web apps with REST APIs and React frontends.
- Improved reliability by diagnosing and resolving production issues during on-call rotations.
- Built backend services in C#/.NET for production systems.
- Delivered standalone SaaS components and improved performance through debugging and optimization.