search menu icon-carat-right cmu-wordmark
quotes
2023 Year in Review

CMU Collaborations Enhance Outcomes for U.S. Government

Tackling the nation’s toughest challenges in software engineering, cybersecurity, and artificial intelligence requires a rare breadth and depth of knowledge. As an integral part of Carnegie Mellon University (CMU), the SEI has access to a talented and experienced pool of domain experts. Research and development for the Department of Defense and other U.S. government agencies was enhanced by the SEI’s collaboration with CMU faculty, staff, and students on the following projects in 2023:

  • Automatic Detection of Malicious Code—Ruben Martins, assistant research professor in the Computer Science Department, developed large language model approaches to automatically determine if system application programming interface (API) functions could leak sensitive information, thereby easing the burden on security analysts.
     
  • Building a Security Operations Center (SOC) Knowledge Base and Ontology—Travis Breaux, associate professor of computer science and director of the Requirements Engineering Lab, guided the development of a first-of-its-kind ontology of SOC expert knowledge and the description logic to turn that ontology into a functioning tool for operationalizing new SOC capabilities.
     
  • Risk-Aware Adaptive Moving Target Defense (MTD)—Ehab Al-Shaer, distinguished career professor in the Software and Societal Systems Department, investigated appropriate risk metrics and defense vectors for a capability to defend systems against advanced persistent threats proactively.
     
  • Portable High-Performance Inference on the Tactical Edge (PHITE)Tze Meng Low, associate research professor with the Department of Electrical and Computer Engineering, and his graduate students Upasana Sridhar, Nicholai Tukanov, and Elliott Binder, helped develop Software for Machine Learning Libraries (SMaLL), an open source machine learning (ML) framework for low-power devices. The PHITE project enables tactical edge devices to use ML more efficiently.
     
  • Automatic Detection of Stakeholder Assumption Mismatches in ML System DevelopmentChristian Kästner, associate professor and director of the software engineering PhD program, and his doctoral students Nadia Nahar and Chenyang Yang, helped identify stakeholder collaboration challenges and tools to test model production readiness.
     
  • 2022 President’s Cup Cybersecurity CompetitionFaculty and students of the Center for Transformational Play (CTP) and Entertainment Technology Center (ETC) developed the video game Cubespace, based on challenges designed by the SEI, for the finals of the Cybersecurity and Infrastructure Security Agency’s annual contest for federal cybersecurity practitioners.