Panagiotis Kalnis

Also known as Panos Kalnis

Panagiotis Kalnis is a computer scientist at KAUST whose systems research covers large graphs, parallel computing and machine learning on distributed hardware. A University of Patras graduate in computer engineering, he earned his PhD at the Hong Kong University of Science and Technology in 2002. In 2013 he coauthored Mizan, a Pregel graph-processing system, with IBM colleagues, and from 2014 to 2018 he chaired KAUST's Computer Science program. His team built ScaleMine, a frequent subgraph mining system ten times faster than earlier methods, in 2016, and went on to a deep learning model for gene-disease patterns in 2019 and layer-by-layer compression for distributed deep learning in 2020. He is a professor of computer science at KAUST and principal investigator of its InfoCloud research group.

Now at: King Abdullah University of Science and Technology

Timeline

1998–2002

Patras diploma, HKUST doctorate

He receives a diploma in computer engineering from the University of Patras in 1998. In 2002 he completes a PhD at the Hong Kong University of Science and Technology, supervised by Dimitris Papadias.
By 2009

Data anonymization and Stanford

He spends a 2009 sabbatical at Stanford University after serving as an assistant professor at the National University of Singapore. In 2009, with Gabriel Ghinita, Panagiotis Karras and Nikos Mamoulis, he creates heuristics that solve the problems of the k-anonymity and l-diversity data privacy models in linear time.
2013

The Mizan graph system

In 2013, he and colleagues from IBM's Thomas J. Watson Research Center develop Mizan, a Pregel graph-processing system that balances load by migrating vertices at runtime. The paper appears at EuroSys 2013 in Prague and reports up to 84% improvement over static graph pre-partitioning on highly dynamic workloads.
2014–2018

Chairing KAUST computer science

He chairs the Computer Science program at KAUST from 2014 to 2018. In December 2014, as an associate professor leading the InfoCloud group, he gives a talk at the University of Patras on big data on large computers with applications in bioinformatics.
2016

ScaleMine for subgraph mining

In 2016, he leads a team at KAUST's Extreme Computing Research Center that develops ScaleMine, a massively parallel system for frequent subgraph mining in large graphs. ScaleMine runs ten times faster than existing methods.
2019

Deep learning for gene-disease patterns

In 2019, he and computational bioscientist Xin Gao develop a deep learning model at KAUST that improves the ability to identify genes potentially involved in disease.
2020

Compression for distributed deep learning

In 2020, he and postdoctoral research fellow Aritra Dutta work on compressed communication for distributed deep learning in KAUST's Extreme Computing Research Center. Their layer-based approach raises the efficiency of training artificial intelligence models.
As of October 2026

Professor and InfoCloud leader

He is a professor of computer science at KAUST and principal investigator of its InfoCloud research group. His research interests include large graphs and systems for machine learning.