Disputation: Karthik Shivashankar

Doctoral candidate Karthik Shivashankar at the Department of Informatics, Faculty of Mathematics and Natural Sciences, is defending the thesis The Dual Role of Machine Learning in Technical Debt Management: Applying Machine Learning for Identification While Analyzing Debt within ML Systems  for the degree of Philosophiae Doctor.

Tid og sted: , Kristen Nygaards sal (5370), Ole Johan Dahls hus/ Zoom
Bildet kan inneholde: briller, hår, smil, briller, hake.

Private

The PhD defence will be partially digital, in Kristen Nygaards sal (5370), Ole-Johan Dahls hus and streamed directly using Zoom. The host of the session will moderate the technicalities while the chair of the defence will moderate the disputation.

Ex auditorio questions: the chair of the defence will invite the attending audience at Kristen Nygaards sal to ask ex auditorio questions. 

Trial lecture

"Using Large Language Models to Identify Security Vulnerability"

Time and place: October 6, 2025 11:15 AM. Kristen Nygaards sal (5370), Ole-Johan Dahls hus/ Zoom

 

Main research findings

Modern society runs on complex software, and we are increasingly relying on Artificial Intelligence to power everything from healthcare to communication. But this complexity creates hidden problems, known as "technical debt"—shortcuts and design flaws that make software fragile, insecure, and expensive to maintain. The specialized tools needed to find these problems have not kept up, especially for the unique challenges within AI systems themselves. Therefore, my research developed a new generation of AI-powered tools to automatically manage this debt. I created systems that use natural language processing to understand developer discussions and identify quality issues before they escalate. I also built a specialized static analyzer, "MLScent," that acts like a quality inspector for AI code, detecting unique flaws in popular frameworks like PyTorch and TensorFlow. The impact of this work is more reliable and maintainable software, ensuring the technology we depend on is built on a solid foundation.

Adjudication committee

  • Principal researcher Ipek Ozkaya, Carnagie Mellon University Software Engineering Institute (SEI), USA
  • Professor Jingyue Li, NTNU, Norway

  • Associate Professor Michael Kirkedal Thomsen, Department of Informatics, University of Oslo, Norway

Supervisors

  • Professor Antonio Martini, University of Oslo, Norway

  • Professor Geir Kjetil Sandve, University of Oslo, Norway

Chair of defence

  • Professor Viktoria Stray, Department of Informatics, University of Oslo

Contact information at Department: Mozhdeh Sheibani Harat

Publisert 22. sep. 2025 08:30 - Sist endret 1. okt. 2025 08:55
Del på e-post