Alexandros (Alex) Haridis is a registered architect (M.Arch) and design computing researcher with a Ph.D. and M.S. in Architecture: Design and Computation from the Massachusetts Institute of Technology. His research focuses on architectural intelligence, architectural criticism, design theory and methods, and artificial intelligence. As an author of publications, exhibitions, and software artifacts, he examines the co-evolution of architecture and design with digital technologies and formal computation. In Fall 2026, he is a visiting scholar at the University of Stuttgart’s Cluster of Excellence Institute for Computational Design and Construction (IntCDC) and a Research Affiliate at the MIT Department of Architecture. He has held a faculty appointment at Harvard University and collaborates as a design computing expert with organizations in the built environment, education, and design technology sectors.

His current research addresses AI systems that compare and rank architecture and design. He develops design methods that make the judgments encoded in these systems visible, explainable, and open to challenge. Most recently he has pursued this through Beyond Data-Driven Aesthetics, a research platform he directs. Its eponymous research exhibition was presented at MIT’s Keller Gallery (April–June 2026) and featured in MIT News, Kathimerini, AI Art Magazine, and other outlets.

A related strand of his research, Machine Learning for Design and Cultural Collections, treats architectural and design archives as testbeds for machine learning. Collections of drawings and other visual media have long recorded and transmitted design knowledge. His studies ask how they can also serve as computational resources for AI systems, and they emphasize carefully curated, theory-informed datasets over large-scale data collection. Outputs include the JONES-19 dataset, a cultural design dataset downloaded more than 2,000 times, and articles published in Design Computing and Cognition and ACM’s Culture and Computer Science.

These questions build on a foundation in the mathematics of shape and design computation. His journal articles ask how shapes, and rule-based operations on them, can give a formal computational description of design reasoning in CAD systems, supporting more robust representations in practical design workflows. This work contributes new methods to computability theory in shape grammars and to structural mathematics (topology) of shapes. It draws its problems and empirical grounding from architecture and related spatial practices. It has appeared in Environment and Planning B (2022), Computers & Graphics (2022), the Journal of Mathematics and the Arts (2020, 2026), and Nexus Network Journal (2026).

His dissertation, Visual Calculating Aesthetic Value: Formal Models of Description and Evaluation, extends this work into a uniform computational framework for encoding aesthetic, spatial, and disciplinary judgments traditionally exercised by design professionals. Formalizations of these judgments had previously been dispersed across distinct twentieth-century computing paradigms.

Haridis has received fellowships, grants, and awards from the MIT School of Architecture and Planning, the Harvard Data Science Initiative, and the A.G. Leventis and Onassis Foundations, among others. He has presented his work by invitation at Dartmouth College, MIT, the University of Pennsylvania, Columbia University, the New York Institute of Technology, and the National Technical University of Athens.

While Haridis’s research is anchored in design computing and architecture, his teaching bridges studio pedagogy and engineering problem-solving. He has taught graduate and undergraduate courses and studios in architecture, design engineering, visualization technologies, design methods, and human values in technology. These include courses in Harvard’s Master in Design Engineering program and courses offered jointly by the Graduate School of Design and the School of Engineering and Applied Sciences.

Current Interests/Active Questions