Jendrik Poloczek

Scala Backend Developer

Berlin, Germany

Hi, I’m Jendrik Poloczek, a data engineer in Berlin, Germany. My interests related to development are: innovative and concise code in Scala and Python. I have a machine learning background and also have real-world experience in the field of big data. Besides, I love to drink good tea and travel.

Work Experience

  • Hypercapitalistic

    Founder |Jul, 2017Current

    Hypercapitalistic is an AI-powered real-time news aggregation for cryptocurrency from over hundreds of sources including popular online magazines, Twitter, YouTube and podcasts.

    • Developed a microservice-oriented Scala backend with technologies like Twitter's Finagle, AWS EC2, AWS CodeDeploy and DynamoDB
    • Developed a ReactJS-based frontend with GrommetUX
    • Designed logo, corperate identity and UX
  • Hive Streaming AB

    Scala Backend Engineer|Apr, 2016Apr, 2017

    I was full-time employed as a backend engineer in the field of P2P video distribution. The team consisted of only two backend developers in total. The work involved architectural design of distributed systems and development with technologies like Scala, Finagle, Azure, Docker, Kafka, Kafka Streams & Spark for the analytics part.

    • Event-Time data processing with Kafka Streams including writing of Kafka Streams topologies and APIs for accessing results
    • Batch Job / Batch Data API for triggering and managing Spark processes and retrieval of results
    • Developed the mockedstreams framework for testing Kafka Streams topologies without Kafka and Zookeeper
  • plista GmbH

    Data Engineer|Oct, 2014Nov, 2015

    I am full-time employed as a data engineer in the field of article recommendations for online publishers. This work involves machine learning and text mining skills, as well as designing software architectures with Scala, Akka and Apache Spark.

    • Developed a streaming recommender for news articles in Akka and Spark Streaming
    • Introduced Kafka and replaced RabbitMQ as message broker
    • Introduced Spark and Spark streaming into the company
    • Product-owner Onsite Recommendations
  • Carl von Ossietzky Universität Oldenburg

    PHD Researcher|20132014

    After graduation with a master’s degree in Computer Science at the Carl-von-Ossietzky University in Oldenburg, Germany in 2013, I worked as a researcher in the Computational Intelligence group. Besides data mining and machine learning in the field of wind energy analysis and prediction, I gained experience in algorithm design and empirical experiments. Furthermore, I worked on WindML, an open source framework in Python for wind energy forecasting with machine learning approaches, and on MetaOpt, a library that optimizes black-box functions using a limited amount of time and utilizing multiple processors.

    • Jendrik Poloczek, Oliver Kramer: Multi-Stage Constraint Surrogate Models for Evolution Strategies. KI 2014. Springer
    • Jendrik Poloczek, Nils André Treiber, Oliver Kramer: KNN Regression as Geo-Imputation Method for Spatio-Temporal Wind Data. SOCO 2014. Springer
    • Jendrik Poloczek, Oliver Kramer: Local SVM Constraint Surrogate Models for Self-adaptive Evolution Strategies. KI 2013: 164-175 Springer
    • Oliver Kramer, Fabian Gieseke, Justin Heinermann, Jendrik Poloczek, Nils André Treiber: A Framework for Data Mining in Wind Power Time Series. DARE 2014.

Education

  • Carl von Ossietzky Universität Oldenburg

    Master of Science - Computer Science|20072013

    1.2

Projects

  • Mocked Streams

    Creator|2017Current

    Mocked Streams is a library for Scala which allows you to unit-test processing topologies of Kafka Streams applications without Zookeeper and Kafka Brokers.

  • Bottle Dynamo

    Creator|Jun, 2017Current

    Good enough AWS DynamoDB abstraction in Scala with Circe JSON serialization using Twitter Futures.

  • WindML and MetaOpt

    Creator|20152016

    WindML is an open source framework in Python for wind energy forecasting with machine learning approaches, and on MetaOpt, a library that optimizes black-box functions using a limited amount of time and utilizing multiple processors.

  • 8bitdash

    Creator|2015Current

    8bitdash is a curated collection of pixel art pieces as start page I made in 2015. It was featured on the front page of Reddit and is still a highly-frequently visited site world-wide.

Skills

    • German (native), English (fluent), Italian (basic)
    • Driver's License
    • Leadership
    • Creativity
    • Scala
    • Apache Spark
    • Kafka
    • AWS
    • ReactJS