Biography
Moustafa is a Co-Founder & CTO at Disarray and a CS Researcher at UC Berkeley’s SkyLab. He previously served as a Technical Director at NASA Ames Research Center, where he received NASA’s Early Career Achievement Medal for pioneering secure, real-time AI/ML systems, significantly advancing NASA’s capabilities in space missions and mixed autonomy air traffic management. Prior to NASA, he was an EECS postdoc at UC Berkeley’s RISELab under the guidance of David E. Culler.
Moustafa earned his Ph.D. from Rutgers under the guidance of Manish Parashar, where he was a three-time recipient of the IBM Ph.D. Fellowship, won the IEEE IPDPS Ph.D. Forum Award, and received the Rutgers ECE Academic Achievement Award. His research has also been recognized with the IEEE/ACM CCGrid Scale Challenge Award, the ACM UCC Cloud Challenge Award, and was featured in Fortune magazine.
Current Focus
Moustafa’s current research focuses on compound AI systems and on grounding autonomous machine-learning agents in data semantics and organizational context, addressing core limitations in agent autonomy, reliability, and evaluation that arise when agents operate without a faithful understanding of real-world data, infrastructure, and system dependencies.
At Disarray, Moustafa leads research on knowledge graph construction from heterogeneous data assets and source code. The approach uses static program analysis to identify data dependencies, semantic embeddings for context retrieval, and graph-based provenance tracking to enable agents to reason about actual system state rather than hallucinated abstractions. This differs from standard RAG approaches through explicit dependency modeling, provenance tracking, and structural understanding of code-data relationships.
Research Agenda
- Provenance-aware code generation: Investigating how knowledge graphs enable agents to generate code with explicit lineage tracking, enabling auditability and debugging
- Context management at scale: Addressing systems challenges of maintaining real-time knowledge graphs across heterogeneous data stores
- Robustness through semantic grounding: Understanding which agent failure modes can be mitigated through structured context