Anastasia Mavridou
Senior Research Scientist
KBR Inc / NASA Ames Research Center
Technical Lead: NASA's FRET project
I like turning “it should probably work” into “it shall, provably.” 🚀
Do you speak FRETish?
FRETish requirement
In visitor mode, upon your arrival at this page, the site shall immediately show you how Anastasia turns ambiguous requirements into specifications an autonomous system can be verified against.
Click any colored word below and type your own
In visitor mode, upon you scroll to this section, this widget shall immediately let you write your own FRETish requirement.
about
I'm the Technical Lead of FRET, NASA's open-source tool that turns ambiguous, natural-language requirements into precise specifications a machine can check, and I lead the team of engineers and researchers who build it. I'm also a member of the Robust Software Engineering (RSE) Group at NASA Ames Research Center, where I research and build tools for verifying and validating safety-critical systems.
In collaboration with engineers, I apply FRET to help verify complex, often autonomous, aerospace systems before they fly.
I completed my postdoctoral research at Vanderbilt University's Institute for Software Integrated Systems, working under Prof. Janos Sztipanovits. I hold a Doctor of Science from École Polytechnique Fédérale de Lausanne (EPFL), where I was fortunate to be advised by Prof. Joseph Sifakis (2007 Turing Award laureate) and Prof. Simon Bliudze.
news
Delivered the keynote “Fighting AI with AI: A Requirements-based Approach” at the Software Engineering and AI Symposium, Hasso Plattner Institute, Germany, on using foundation models to assure AI-enabled systems starting from formal requirements. The ideas behind the talk are laid out in “Fighting AI with AI: Leveraging Foundation Models for Assuring AI-Enabled Safety-Critical Systems”.
Multiple papers accepted this year at conferences including ICSE, NFM, FM, and FormaliSE: “Automating Requirements Formalization: Using LLMs and Low-Complexity Distinguishing Traces for Semantic Validation”, “Resource-Constrained Systems: Integrating FRET and R2U2”, running FRET alongside R2U2, and “Requirements Elicitation, Formalization, and Analysis with FRET: A Tutorial”.
“A Taxonomy of Human-Robot Teamwork Requirements”, co-authored with Hazel Taylor, Sandy Lozito, Louise Dennis, Michael Fisher, and Marie Farrell, accepted at RE 2026, proposing a taxonomy for classifying requirements around how humans and robots coordinate as a team, not just how a single robot behaves on its own.
Co-edited the proceedings of FormaliSE 2025, the International Conference on Formal Methods in Software Engineering, for Springer, with Gwen Salaün.
Gave invited lectures on FRET at SPACERAISE, an international doctoral school in L’Aquila, Italy, training the next generation of space software engineers in AI, robotics, and digital twins for the space sector.
career
Senior Research Scientist, KBR Inc. at NASA Ames Research Center. I'm the Technical Lead of FRET, a NASA open-source tool that turns natural-language requirements into precise, formally analyzable specifications.
Research Scientist, KBR Inc. at NASA Ames Research Center. I worked on connecting FRET to downstream analysis tools for model checking and runtime monitoring.
Postdoctoral Scholar, Vanderbilt University, Institute for Software Integrated Systems. I worked with Prof. Janos Sztipanovits, returning to a lab I'd first visited as a scholar in the summer of 2015.
Doctor of Science in Computer Science, EPFL. My thesis, “Modelling Architecture Styles,” was advised by Prof. Joseph Sifakis (2007 Turing Award) and Prof. Simon Bliudze. I spent these years as a Research Assistant in the Rigorous System Design Laboratory.
M.Sc. in Computer Science, University of Tulsa. Thesis: “A Situational Awareness Framework for the Smart Grid,” advised by Prof. Mauricio Papa.
Diploma in Electrical & Computer Engineering, Aristotle University of Thessaloniki. A 5-year degree; my thesis covered information assurance for cybersecurity and critical-infrastructure protection.
selected publications
Generation of Formal Requirements from Structured Natural LanguageHonorary mention: top 3 papers