Enes Zvornicanin, Developer in Sarajevo, Federation of Bosnia and Herzegovina, Bosnia and Herzegovina
Enes is currently unavailable

Enes Zvornicanin

Data Scientist and Developer

Sarajevo, Federation of Bosnia and Herzegovina, Bosnia and Herzegovina

Toptal member since August 16, 2021

Bio

Enes is a data scientist with over six years of experience, including five years working remotely with US-based teams from The World Bank and a Silicon Valley startup, Yewno. He focuses on developing state-of-the-art machine learning solutions using time series, tabular, and text data. Enes has experience working on end-to-end ML and data-based projects, including analysis, research, development, and deployment. He is familiar with MLOps and data engineering best practices and tools.

Portfolio

World Bank Group
Python, Pandas, Polars, NumPy, NetworkX, Google Maps API, Bing Maps API...
Self-employed
Machine Learning, Data Science, Natural Language Processing (NLP), Gensim...
Tech 387
Python, Data Visualization, Data Analysis, Business Analysis, Pandas, Plotly...

Experience

  • Python - 6 years
  • Pandas - 5 years
  • Scikit-learn - 5 years
  • Data Analysis - 5 years
  • Machine Learning - 5 years
  • Data Science - 5 years
  • Natural Language Processing (NLP) - 4 years
  • Time Series - 3 years

Preferred Environment

Slack, Jira, Google Hangouts, Python, Amazon Web Services (AWS), Google Cloud

The most amazing...

...thing I've developed is a text cleaning and entity linking system for messy firm names, which the World Bank uses to handle millions of records.

Work Experience

Data Science Consultant

2022 - PRESENT
World Bank Group
  • Worked remotely as a data science consultant on an NLP project with a team based in Washington, DC.
  • Developed and optimized a complex system for text cleaning, named-entity resolution, data clustering, and entity linking on very messy text data.
  • Worked with tabular geospatial data with a few billion rows. Developed multiple Python scripts to extract and validate geo-specific terms using Google Maps and Bing Maps API.
Technologies: Python, Pandas, Polars, NumPy, NetworkX, Google Maps API, Bing Maps API, Natural Language Processing (NLP), Fuzzy Logic, Unsupervised Learning, Big Data, Slack, Geospatial Data, Machine Learning, Data Science, OpenAI GPT-4 API, Gemini API, ChatGPT

Freelance Data Scientist | Technical Writer

2021 - PRESENT
Self-employed
  • Developed several biomedical NLP pipelines using Gensim Word2Vec and BERT-type models. Worked on tasks related to NER and topic modeling.
  • Built a machine-learning credit score system focused on the interpretability of the models. Developed and deployed a Python script for crypto trading using the Binance API.
  • Worked on several projects related to time-series forecasting, speech-to-text, summarization, and YouTube video automation.
  • Wrote more than 50 articles and tutorials related to ML and data science topics available at eneszvornicanin.com.
Technologies: Machine Learning, Data Science, Natural Language Processing (NLP), Gensim, Hugging Face, Pandas, StatsModels, ARIMA, Keras, TensorFlow, PyTorch, Python, Jupyter Notebook, Git, SQL, SQLite, Scikit-learn, SHAP, XGBoost, Binance API, Amazon EC2, Amazon S3 (AWS S3), Unix, NumPy, Data Analysis, Statistical Modeling, Deep Learning, Artificial Intelligence (AI), Data Engineering, OpenAI GPT-4 API, Gemini API, LangChain, Whisper, YouTube API, Azure Functions, ChatGPT, Large Language Models (LLMs), Optical Character Recognition (OCR), AI Content Creation, Data Visualization, Flask, Dash

Senior Data Scientist

2021 - 2022
Tech 387
  • Analyzed data from a mobile app and extracted business metrics, including conversion rate, CPU, user engagement, user segmentation, and survival analysis. Identified possible issues and recommended a business strategy for future development.
  • Developed and deployed a system that helps project managers to track Jira tickets. The system uses data from Jira API, sends notifications and alerts through Slack bots and email, and shows charts on a website that is deployed on AWS using Flask.
  • Developed web scraping Python scripts and organized internal data in the company.
  • Created several Slack bots and ETL processes using Python.
Technologies: Python, Data Visualization, Data Analysis, Business Analysis, Pandas, Plotly, Flask, Matplotlib, Mixpanel API, Google Cloud Platform (GCP), Amazon EC2, Jira REST API, Slack API, Firebase, Git, Cron, Jupyter Notebook, Selenium, SQL, Data Analytics, Data Science

Lead Data Scientist

2020 - 2021
Entropy387 (Yewno)
  • Developed and trained over one million diverse machine learning models for stock market movement prediction, which improved the trading strategy by approximately 30% in cumulative return.
  • Applied data preprocessing, feature engineering, and optimization techniques for tuning ML models.
  • Collaborated closely with high-level colleagues from finance and business to brainstorm ideas for new products and improvements.
Technologies: Machine Learning, Artificial Intelligence (AI), Python, Time Series, Data Science, Classification, Regression, Optimization, Data Cleaning, Pandas, Scikit-learn, Keras, NumPy, Stock Analysis, Jira, Git, Econometrics, Networks, Matplotlib, Data Analysis, Modeling, Amazon Web Services (AWS), Jupyter, Amazon S3 (AWS S3), Data Analytics, Trading, Statistical Modeling, Microsoft Excel, Jupyter Notebook, Natural Language Processing (NLP), Deep Learning, Neural Networks, Statistics, XGBoost, TensorFlow, Slack, Google Hangouts, Linux, PyCharm, ETL, Algorithms, Data Visualization

Data Scientist

2019 - 2020
Entropy387 (Yewno)
  • Conducted research and developed several state-of-the-art solutions using graph embedding techniques, network anomaly detection methods, and reinforcement learning.
  • Prepared and held presentations regarding data QA, cleaning, and analysis using Jupyter notebooks.
  • Built several ETLs and collaborated closely with the data engineering team preparing data pipelines for production.
Technologies: Python, Machine Learning, Optimization, Artificial Intelligence (AI), Algorithms, Programming, SQL, Time Series, Data Science, Classification, Data Cleaning, Pandas, Scikit-learn, Keras, NumPy, Clustering, Stock Analysis, Jira, Git, Networks, Matplotlib, Data Analysis, Jupyter Notebook, Slack, Google Hangouts, Linux, PyCharm, Modeling, Amazon Web Services (AWS), Jupyter, Amazon S3 (AWS S3), Data Analytics, Microsoft Excel, ETL, Data Visualization

Data Scientist

2018 - 2019
Cape Ann Enterprise
  • Researched and developed several neural networks and machine learning methods for classification.
  • Implemented a pipeline for NLP tasks similar to sentiment analysis.
  • Prepared and cleaned data for classification tasks.
Technologies: Python, Jupyter Notebook, Keras, Pandas, Natural Language Processing (NLP), Data Science, Machine Learning, Deep Learning, Neural Networks, Modeling, Jupyter, Data Analysis, Microsoft Excel, Artificial Intelligence (AI), Programming, Classification, Data Cleaning, Matplotlib, Linux, PyCharm, Data Analytics, Algorithms

Experience

Invest Info Web App

https://investba.info/
A Python-based web app that analyzes the state of the real estate market in Bosnia, along with stock options, cryptocurrency, and economic data. It provides graphical insights into real estate price trends, some stock options graphs, cryptocurrency charts, and economic data.

Building MLOps Pipeline for Time Series Prediction [Tutorial]

This tutorial describes and implements an end-to-end time-series project based on the Binance trading app, utilizing MLOps architecture. The project follows best CI/CD practices and incorporates technologies like GitHub Actions, Docker, Amazon ECR, ECS, EC2, S3, Neptune API, XGBoost, Optuna, cron jobs, and more.

Overview of Time-series Forecasting Methods

https://www.kaggle.com/eneszvo/time-series-forecasting-p1-es-arima-var
The purpose of this project is to provide a simple and clear theoretical explanation and minimal working examples of several models for time series forecasting from econometrics. Besides that, in the Jupyter notebook are explained some terms such as stationarity, ACF, PACF, and so on.

Automated Testing in Machine Learning Projects [Best Practices for MLOps]

The article on automated testing in machine learning explains the types of automated tests, including smoke testing, unit testing, integration testing, regression testing, data testing, model testing, and monitoring testing.

Topic Modeling and Latent Dirichlet Allocation (LDA)

https://datascienceplus.com/topic-modeling-and-latent-dirichlet-allocation-lda/
This article explains the natural language processing technique for topic modeling using latent Dirichlet allocation (LDA). Besides the theoretical part, there is a simple tutorial in Python with some NLP preprocessing techniques.

Shopee Summary—Matching Products Using Images and Titles

https://www.kaggle.com/eneszvo/shopee-summary-efficientnet-arcface-bert
This project is a summary of the "Shopee—Price Match Guarantee" competition where the main goal was to match the same products based on their images and titles. In this, a Jupyter notebook presented an overview of different methods for detect similar images or sentences using pre-trained neural network embeddings, pHash, TF-IDF, and similar.

Time Series Projects: Tools, Packages, and Libraries That Can Help

This article presents the most popular Python packages and libraries that can be used in any time-series project. Also, the concept of time series and some examples are introduced. In the end, is presented a comparison of these packages.

Introduction to Crypto Bitcoin Trading with Python and Binance

This article explains how to download historical data from Binance, build a simple trading strategy with Python, backtest and optimize it, and finally deploy it on AWS. All code is shared in the GitHub repository.

Education

2018 - 2020

Master's Degree in Computer Science

University of Sarajevo, Faculty of Natural Sciences and Mathematics - Sarajevo, Bosnia and Herzegovina

2014 - 2018

Bachelor's Degree in Applied Mathematics

University of Tuzla, Faculty of Natural Sciences and Mathematics - Tuzla, Bosnia and Herzegovina

Skills

Libraries/APIs

Pandas, Scikit-learn, Matplotlib, NumPy, XGBoost, Keras, TensorFlow, Mixpanel API, Jira REST API, Slack API, PyTorch, Binance API, NetworkX, Google Maps API, Bing Maps API, YouTube API

Tools

PyCharm, Jupyter, Jira, Slack, Google Hangouts, Microsoft Excel, ChatGPT, Git, Plotly, Cron, Gensim, StatsModels, ARIMA, Whisper

Languages

Python, SQL

Platforms

Jupyter Notebook, Linux, Amazon Web Services (AWS), Google Cloud Platform (GCP), Amazon EC2, Firebase, Unix, Visual Studio Code (VS Code), Docker, Azure Functions

Paradigms

ETL

Storage

Amazon S3 (AWS S3), SQLite, Google Cloud

Frameworks

Flask, Selenium

Other

Machine Learning, Optimization, Artificial Intelligence (AI), Applied Mathematics, Programming, Time Series, Data Science, Classification, Data Cleaning, Data Analysis, Natural Language Processing (NLP), Statistics, Modeling, Data Analytics, Data Visualization, Algorithms, Regression, Clustering, Stock Analysis, Econometrics, Networks, Deep Learning, Neural Networks, Statistical Modeling, Unsupervised Learning, Optical Character Recognition (OCR), BERT, Image Processing, Trading, Object Detection, Video Processing, Business Analysis, Hugging Face, SHAP, Data Engineering, Machine Learning Operations (MLOps), Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), Polars, Fuzzy Logic, Big Data, Geospatial Data, CI/CD Pipelines, Amazon Neptune, Deployment, OpenAI GPT-4 API, Gemini API, LangChain, Large Language Models (LLMs), AI Content Creation, Dash, Cursor AI, Data Scraping

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring