Milos Grubjesic, Software Developer & Data Engineer Developer in Novi Sad, Vojvodina, Serbia
Milos Grubjesic

Software Developer & Data Engineer Developer in Novi Sad, Vojvodina, Serbia

Member since April 19, 2015
Milos is a senior software developer and data scientist with over a decade and a half of education and experience using machine learning and AI, Python, R, Scala, and more to solve complex research problems and enrich the capabilities of web and software projects.
Milos is now available for hire

Portfolio

Experience

Location

Novi Sad, Vojvodina, Serbia

Availability

Part-time

Preferred Environment

Ubuntu, RStudio, PyCharm, IntelliJ IDEA, Jupyter

The most amazing...

...thing I've implemented is a fake user classifier—an ensemble of several with optimized coefficients, actually—to help remove bad guys from a dating site.

Employment

  • Data Scientist

    2010 - PRESENT
    Freelance
    • Created classifications (e.g., fake user classifier).
    • Built regression models (e.g., price predictions).
    • Developed recommendation engines (PredictionIO, Mahout, both collaborative filtering and content-based filtering).
    • Created indices (artist's index, popularity index, demand index, etc.).
    • Developed time series analysis and forecasting.
    • Used NLP, specifically bag of words and Word2vec (document similarity, recipes parsing).
    • Created predictive analytics (future contracts analysis).
    • Analyzed statistics using Bayesian inference (tennis match result predictions).
    • Wrote parallel algorithms (speeding up an application for Ford based on CUDA).
    • Forecasted a linear TV viewership time series.
    Technologies: Keras, TensorFlow, Apache Spark, R, Python
  • Researcher

    2009 - 2010
    Yokohama National University, Japan
    • Researched fundamentals of real-world databases, including NoSQL databases for PET scannner images.
    • Cracked Java licenses by using ASM, byte code manipulation framework, and aspect-oriented programming.
    Technologies: NoSQL, Java
  • Java Software Developer

    2008 - 2009
    Levi9
    • Created a Java web application for a Dutch customer.
    • Worked on a site in Amsterdam for one client.
    Technologies: Spring, Java
  • Java Software Developer

    2006 - 2008
    Zesium
    • Implemented Java web applications.
    • Maintained Linux machines as a junior admin.
    • Conducted white hat testing of the web application.
    • Worked on complex Java application for health diseases for Danish customers.
    • Tested the application under a high load of requests.
    Technologies: Linux, Spring, Java

Experience

  • Data Pipelines with Apache Spark and Scala (Development)

    Implemented complex data pipelines on an Apache Spark cluster. Cleaned, deduplicated, and aggregated data using Scala and Apache Spark.

  • Scala Implementation of Calculator (Development)
    https://github.com/milosjava/calculator

    Simple scientific calculator app that supports basic arithmetic operators as well as log on floating point numbers. It can solve simple linear equations with a single variable and parentheses, and has a language parser.

  • Algorithms Time Complexity in Scala (Other amazing things)
    https://github.com/milosjava/algorithms-time-complexity

    Code exercises in Scala regarding different algorithms with time and complexity constraints.

  • Predictive Analytics (Development)

    A project for commodity market predictive analytics. Includes but not limited to time series forecasting and statistical measures/analysis. Implemented various classifiers/regressions for identifying best trading strategies, back-testing, etc.

    Technologies: R, Apache Spark, AWS EC2
    Role: Data Scientist (1 year)

  • Price Predictions (Development)

    An exciting site for locking computer games prices in the future. Implemented a price prediction system based on historic data from various sources (e.g., Amazon). Also, implemented various indexes including popularity/demand indexes.

    Technologies: R , MySQL
    Role: Data Scientist (8 months)

  • Online Dating Site (Development)

    Worked as data scientist on recommendation engine for user recommendations, fake/scam integrated user classification, user clustering, preference prediction, FIM on user features, and many more.

    Technologies: Java, Spring, Mahout, Hadoop, WEKA, Pentaho Data Integration (Kettle), MongoDB, MySQL, and more
    Role: Data Scientist (1 year)

  • Look a Like Modeling (Development)

    A project involving several sub systems: data tidying and checking, exploratory analysis, dimensionality reduction, and prediction modeling.

    Technologies: R, Apache Spark
    Role: Data Scientist (6 months)

  • Recipes Web Application (Development)

    Implemented recommendation engine based on collaborative filtering and content-based filtering. Also did all the math and programming for a flavor-match algorithm. Fixed and cleaned NLP results for ingredient types and quantity recognition. Built up a Java (Spring) based WS that handles all requests for recommendations and flavors.

    Technologies: Java, Hadoop , Spring, Mahout, PDI , WEKA, MySQL, and more
    Role: Data Scientist (7 months)

Skills

  • Languages

    R, Python, Python 3, Scala, SQL, Java
  • Frameworks

    Apache Spark, Spark, Flask, Spring
  • Libraries/APIs

    Pandas, NumPy, SciPy, Sklearn, Scikit-learn, Caret, Keras, TensorFlow, Spark ML
  • Tools

    PyCharm, IntelliJ IDEA, Spark SQL, Jupyter, Weka, Pentaho Data Integration (Kettle), Git
  • Platforms

    RStudio, Ubuntu Linux, Jupyter Notebook, Linux, Ubuntu, Azure, Databricks, Amazon Web Services (AWS), AWS EC2
  • Other

    Machine Learning, Time Series Analysis, Statistics, APIs, Recommendation Systems, Regression, Classification, Artificial Intelligence (AI), Deep Learning, Predictive Analytics, Data Visualization, AWS, Statistical Data Analysis, Predictive Learning, Predictive modeling
  • Paradigms

    ETL, Data Science, Functional Programming
  • Storage

    NoSQL, PostgreSQL, MySQL, MongoDB

Education

  • Master's degree in Computer Science
    1998 - 2005
    Faculty of Technical Sciences - Novi Sad, Serbia

Certifications

  • Deep Learning Nanodegree
    MARCH 2018 - PRESENT
    Udacity
  • Scalable Machine Learning
    AUGUST 2015 - PRESENT
    EdX
  • Machine Learning
    JULY 2013 - PRESENT
    Coursera

To view more profiles

Join Toptal
Share it with others