Emmanouil Kritharakis

I am a Data Scientist at Satalia, working remotely from Athens, Greece. My research interests are in graph machine learning, systems and near storage computing. Previously, I obtained my master of science from Boston University in the United States and my master of engineering and bachelor's degrees from the Technical University of Crete in Greece.

Email  /  CV  /  LinkedIn  /  Github

profile photo
Research

I am mainly interested in combining machine learning with systems. Much of my research is about inferring graph machine learning algorithms and how can we efficiently work with large enough graphs that do not fit in memory with the help of computational storage devices such as smartSSDs.

Work Experiences
  • Satalia Athens, Greece

    Data Scientist | 02/2024 - Present

  • Boston University Boston, USA

    Research Assistant at CASP system Lab | 09/2021 - 12/2023

    Working under the supervision of Dr. Kalavri on representation learning over large graphs.
    Research interests span over Graph Machine Learning, Streaming Processing Systems, Near Storage Computing.
    Worked as Teaching Assistant for:

    1. CS551 : Streaming and Event-driven Systems (Spring 23’)
    2. CS506 : Data Science Tools & Applications (Fall 23’)

    Extensive use of Python, C++, Pytorch / Libtorch, Cuda, Java, Apache Flink, Apache Kafka, Docker, GCP, AWS.

  • Lawrence Livermore National Laboratory Livermore, USA

    Computing Scolar Research Intern in Computational Mathematics group at CASC | 05/2023 - 08/2023

    Worked under the supervision of Dr. Hill on benchmarking explainability algorithms over Graph Neural Networks (GNNs) to detect malware behavior in software binaries.
    Extensive use of Python, Pytorch, Docker, AWS.

  • Samsung Semiconductor Inc San Jose, USA

    Systems Technology Research Intern at Samsung MSL group | 06/2022 - 09/2022

    Worked under the collaboration of Samsung and CASP Lab academic program, accelerating GNNs training stage thanks to near storage calculation techniques.
    Extensive use of C++, HLS programming for FPGAs, Cuda, OpenCL.

  • Hellenic National Defence General Staff Athens, Greece

    Database Administrator | 09/2020 - 06/2021

    Worked under the ”Unified distributed army mail system” project, by interconnecting the military servers all around Greece to be able to send documents through a web app in the private greek army network.
    Extensive use of PL/SQL for Oracle database and Microsoft SQL Server.

  • LeasePlan Athens, Greece

    Database Software Engineer | 12/2019 - 09/2020

    Worked under the ”Turn Around” project by updating the corporate software with over 40 user-call database features and was responsible for two intercorporate projects:

    1. Estimated rented car’s cumulative kilometers
    2. Built an webservice called E-Acceptance for automated acceptance of new clients
    Extensive use of Python, SQL and RPG language.

  • CERN Organization Geneva, Switzerland

    Summer Research Intern | 06/2019 - 09/2019

    Worked with the EP-SFT group on a generative machine learning model (Variational Autoencoder) to speed-up a high energy physics experimental simulation tool named GEANT4.

Honors and Awards
  • Onassis Foundation Scholarship (2021) for supporting my studies as a master candidate at Boston University Computer Science Department.
  • ANEK Lines Scholarship (2019) for completing my studies in ECE School of Technical University of Crete as a salutatorian in the class of 2014.
  • Excellence Award (2019) for completing my studies in ECE School of Technical University of Crete as a salutatorian in the class of 2014.
Publications
  • Kritharakis, Emmanouil , Shengyao Luo, and Karan Unnikrishnan, Vivek Vombatkere. Detecting trading trends in streaming financial data using Apache Flink. In the 16th ACM International Conference on Distributed and Event-Based Systems, DEBS ’22.
  • Sonia Horchidan, Kritharakis, Emmanouil, Vasiliki Kalavri, and Paris Carbone. Evaluating model serving strategies over streaming data. In the DEEM@ SIGMOD ’22, pages 4–1. (Best Paper Award).
Using template from Jon Barron