"Examination Room Usage Optimization Via Greedy and Exact Methods: A Vietnamese University Case Study," by Hieu Le, Surya Hanjaya, Trang Hoang and Dr. Xuan-Bach Le, has been accepted at SOMET 2026 (International Conference on Intelligent Software Methodologies, Tools, and Techniques) in Kuala Lumpur, Malaysia, September 2026. Most of what it costs a university to run exams comes down to how many rooms have to be opened for each session. Conservative policies and course-specific constraints often leave those rooms poorly used. Unlike examination timetabling work, which optimises time slots and room allocation together, this paper takes on the post-timetabling problem: minimising the rooms needed for already-fixed sessions, subject to room capacity, campus limits and course separation rules. Two cases are examined — improving an existing allocation through post-processing, and allocating rooms from scratch. The approach pairs a fast best-fit decreasing greedy algorithm with an ILP formulation that solves small-to-moderate instances exactly. Experiments on real data from Ho Chi Minh City University of Technology (HCMUT), a multi-campus public university in Vietnam, show substantial room savings and a clear scalability–optimality trade-off. Paper: https://lexuanbach.github.io/publication/SOMET2026.pdf Code: https://github.com/hieukendu/ILP