Resume

My research interests and my experience building large-scale systems.

Experience

• 2025 to now
  AWS Staff Software Engineer

  Amazon EMR, Upper Engine team. Team tech lead for query engine performance and stability.

  Apache Gluten, Velox, Spark, C++, adaptive runtime filtering, query optimization

• 2022 to 2025
  Confluent Staff Software Engineer

  Stream Processing product, Runtime team. Part of the core team that launched a multi-cloud serverless Apache Flink product, from inception to general availability. Led exception classification and the regional artifact service, and worked on Flink user defined functions in Java and Python.

  Apache Flink, Java, Python, multi-cloud stream processing

• 2019 to 2022
  Cloudera Staff Software Engineer

  Data Warehousing product, Runtime team. Open sourced ORC row-level filtering and probe-decode joins, up to 50% faster queries.

  Apache ORC PMC, Hive committer, Tez contributor, columnar storage and query execution

• 2016 and 2017
  Microsoft Research Intern

  Silicon Valley, Cloud and Information Services Lab. Analyzed and optimized scheduling for long-running containers in production clusters.

  Apache YARN, Hadoop contributor, Kubernetes, open source

• 2014 to 2019
  Imperial College London Graduate Researcher

  Department of Computing, Large-Scale Data and Systems group. Built resource management systems for heterogeneous workloads on shared clusters.

  Apache Spark, Flink, Mesos, Hadoop, Kubernetes, cluster resource management

Education

• 2015 to 2019
  PhD Computer Science Imperial College London

  HiPEDS programme. Thesis: Supporting Long-Running Applications in Shared Compute Clusters. Supervised by Peter Pietzuch.

• 2014 to 2015
  M.Res. Advanced Computing Imperial College London

  Distinction.

• 2012 to 2014
  M.Sc. Distributed Systems University of Crete

  First class honours. Supervised by Kostas Magoutis.

• 2007 to 2012
  B.Sc. Computer Science University of Crete

  Ranked 3rd in class.

Papers

• SoCC 2019
  NEPTUNE: Scheduling Suspendable Tasks for Unified Stream/Batch Applications

  Panagiotis Garefalakis, Konstantinos Karanasos, Peter Pietzuch
  ACM Symposium on Cloud Computing, Santa Cruz, California

• SIGMOD 2019
  Peering through the Dark: an Owl's View of Inter-job Dependencies and Jobs' Impact in Shared Clusters

  Andrew Chung, Carlo Curino, Subru Krishnan, Konstantinos Karanasos, Panagiotis Garefalakis, Gregory R. Ganger
  ACM International Conference on Management of Data, Amsterdam · Demo

• EuroSys 2018
  MEDEA: Scheduling of Long Running Applications in Shared Production Clusters

  Panagiotis Garefalakis, Konstantinos Karanasos, Peter Pietzuch, Arun Suresh, Sriram Rao
  13th ACM European Conference on Computer Systems, Porto

• ICDE 2016
  Java2SDG: Stateful Big Data Processing for the Masses

  Raul Castro Fernandez, Panagiotis Garefalakis, Peter Pietzuch
  32nd IEEE International Conference on Data Engineering, Helsinki · Demo

• SRDS 2014
  ACaZoo: A Distributed Key-Value Store based on Replicated LSM-Trees

  Panagiotis Garefalakis, Panagiotis Papadopoulos, Kostas Magoutis
  33rd IEEE Symposium on Reliable Distributed Systems, Nara · Best student paper

• ESOCC 2013
  Towards Cross-layer Monitoring of Multi-Cloud Service-based Applications

  Chrysostomos Zeginis, Kyriakos Kritikos, Panagiotis Garefalakis, Konstantina Konsolaki, Kostas Magoutis, Dimitris Plexousakis
  European Conference on Cloud and Service Computing, Málaga

• DAIS 2013
  Strengthening Consistency in the Cassandra Distributed Key-Value Store

  Panagiotis Garefalakis, Panagiotis Papadopoulos, Ioannis Manousakis, Kostas Magoutis
  13th IFIP Conference on Distributed Applications and Interoperable Systems, Florence

• MultiCloud 2013
  Lifecycle Management of Service-based Applications on Multi-Clouds: A Research Roadmap

  George Baryannis, Panagiotis Garefalakis, Kyriakos Kritikos, Antonis Papaioannou, Kostas Magoutis, Dimitris Plexousakis, Chrysostomos Zeginis
  Workshop on Multi-Cloud Applications and Federated Clouds, Prague

• iThings 2012
  Improving Datacenter Operations Management Using Wireless Sensor Networks

  Panagiotis Garefalakis, Kostas Magoutis
  3rd IEEE International Conference on Internet of Things, Besançon

Selected talks

• 2023
  Flink's pluggable failure handling

  Flink Forward, Seattle, and the Bay Area Apache Flink meetup

  • Flink Forward
  • Meetup
  • Slides
  • FLIP-304
  • Write-up
• 2020
  Accelerating distributed joins in Apache Hive

  ApacheCon

  • Video
  • Slides
  • Track
• 2019
  Cooperative task execution for Apache Spark

  Spark+AI Summit, and SoCC

  • Video
  • Slides
  • Paper
  • SoCC 2019
• 2018
  MEDEA: scheduling long-running applications in shared clusters

  EuroSys, paper

  • Paper
  • Video
  • Slides
  • YARN-6592
• 2017
  Expressive scheduling of long-running containers

  EuroSys, poster, and Hadoop Summit

  • Slides
• 2016
  Java2SDG: stateful big data processing for the masses

  IEEE ICDE, demo

  • Paper
• 2014
  ACaZoo: a distributed key-value store on replicated LSM-trees

  IEEE SRDS, paper, best student paper

  • Paper

Theses

• 2020
  Supporting Long-Running Applications in Shared Compute Clusters

  PhD, Imperial College London

• 2015
  Bridging the Gap between Serving and Analytics in Scalable Web Applications

  M.Res., Imperial College London · slides

• 2014
  A Distributed Key-Value Store based on Replicated LSM-Trees

  M.Sc., University of Crete

Awards

• 2023
  Confluent award, RBAC support delivered under a strict timeline
• 2021
  PMC member, Apache ORC
• 2020
  Committer, Apache Hive
• 2015
  CDT HiPEDS PhD scholarship, Imperial College London
• 2014
  Best student paper, 33rd IEEE Symposium on Reliable Distributed Systems
• 2014
  M.Res. scholarship, Imperial College London
• 2013
  Distinguished teaching assistant scholarship, University of Crete
• 2012
  Graduate research fellow, Institute of Computer Science, FORTH