Shashikant Ilager
Assistant Professor, University of Amsterdam (UvA)
Room L5.21, Lab 42
Science Park 900
1098XH Amsterdam, the Netherlands
s.s.ilager@uva.nl
Hello World!
I am an Assistant Professor at the Informatics Institute (IvI) at the University of Amsterdam (UvA) and a member of the Multiscale Networked Systems (MNS) group. My research lies at the intersection of distributed systems, energy efficiency, and machine learning. I design and optimize high-performance computing platforms, from cloud to edge, with a specific focus on minimizing environmental impact. My current work focuses on the sustainability of data-intensive AI systems, aiming to decarbonize AI infrastructure and enhance energy efficiency without degrading performance.
My research and teaching contributions have been recognized with several honors, including the IEEE TCCLD Outstanding PhD Thesis Award, the IEEE Outstanding Service Award, and Best Paper Awards at CCGRID 2020 and UCC 2023, and Excellence in Teaching Award (UniMelb 2019). My research has been published in leading venues including e-Energy, ASE, ICSOC, EuroPar, and CCGRID, as well as journals such as TPDS, TMC, TC, TSC, TAAS, and TNSM.
Before joining UvA, I held postdoctoral positions at the High Performance Computing (HPC) group at TU Wien, Austria. I completed my PhD in Computer Science and Engineering at the CLOUDS Lab, University of Melbourne, Australia. Additionally, I was a visiting research scientist at IBM T.J. Watson Research Center, USA (Hybrid Cloud Infrastructure team) in Summer 2024 and at IMT Atlantique/INRIA, France, in Winter 2023.
Research Directions
Green AI
Sustainable AI Infrastructure
Measuring, modeling, and reducing the energy footprint of AI workloads — from LLM training and inference to data-intensive pipelines. Includes carbon-aware scheduling, energy-efficient deep learning systems, and sustainability metrics for AI infrastructure.
Cloud Computing
Distributed Systems & Resource Management
Designing and optimizing resource management strategies for large-scale cloud platforms. Covers workload characterization, performance modeling, QoS-aware scheduling, and data-driven approaches to managing heterogeneous cloud infrastructure at scale.
Resource-Efficient Edge
Edge-Cloud Continuum & IoT
Extending intelligence to the network edge — latency-aware task offloading, federated learning under resource constraints, and symbolic/compact AI models for IoT devices. Research spans the full edge-cloud continuum from sensors to data centers.
-
Our paper ABBA-VSM: Time Series Classification using Symbolic Representation on the Edge is accepted at ICSOC 24. Congrats to Meerzhan
.
-
Our paper FLIGAN: Enhancing Federated Learning with Incomplete Data Using GAN is accepted at EdgeSys 24, colocated with EuroSys 24. Congratulations to Paul . Arxiv link
-
Our paper A Data-driven Analysis of a Cloud Data Center: Statistical Characterization of Workload, Energy and Temperature has recieved
Best paper award@ ACM/IEEE UCC 23.