Mojmir Vinkler
Data Mining Developer
Mojmir is a full-stack data scientist with strong development skills, allowing him to handle model design, data collection, and final implementation of software. He has a strong background in statistics, machine learning, business, computer science, and predictive modeling of big data sets.
Portfolio
Experience
Python - 7 yearsData Mining - 5 yearsMachine Learning - 5 yearsDeep Learning - 4 yearsIPython Notebook - 4 yearsApache Spark - 3 yearsPandas - 2 yearsLocation
Availability
Preferred Environment
GitLab, Jupyter, Atom, MacOS
The most amazing...
...thing I've ever built is an optimal product pricing algorithm based on an idea from multi-armed bandits that successfully increased revenues by a third.
Work Experience
Lead Data Scientist
Kiwi.com
- Built a machine learning model from prototyping to deployment.
- Optimized product and business processes.
- Built an interactive web app to visualize A/B testing.
- Set up infrastructure (BigQuery/Spark) to work with petabytes of data.
- Managed and optimized revenue.
Python Programmer, Data Scientist
Human Brain Project
- Built Federated Learning for medical data.
- Implemented custom Federated Learning machine learning algorithms as Docker images.
Python Programmer, Data Scientist
Picwell
- Processed huge amounts of data using Spark.
- Created and optimized scikit-learn models.
- Built an ETL application with Luigi and Flask.
- Created surveys on Mechanical Turk.
- Integrated R and Python code.
Pattern Analysis/Data Matching System Engineer
Flexcode (via Toptal)
- Created an algorithm for predicting correct matches using text analytics and image processing plus a web UI for data collection.
Data Scientist
Webnode
- Handled ETL processing in Python with Pandas.
- Designed and implemented optimal product pricing algorithms.
- Designed a customer lifetime value (CLV) model.
- Monitored AdWords (automated alerts) and bidding based on CLV.
- Engaged in smart business reporting based on statistical models and machine learning.
Web Developer
Mergado.cz
- Created an XML parser for parsing invalid XML files (product feeds for price comparison sites).
- Developed web back-ends.
- Designed and implemented functionality for bidding on price comparison sites.
- Built data connectors to Heureka.cz and Sklik.cz.
- Handled minor front-end work with JavaScript.
Experience
Federated Learning Algorithms
https://github.com/LREN-CHUV/algorithm-repositoryUniversal Portfolios
https://github.com/Marigold/universal-portfoliosSkills
Languages
Python, R, JavaScript, PHP
Frameworks
Apache Spark, GAE, Flask, Spark, Django
Libraries/APIs
Matplotlib, Pandas, PyMC, NumPy, SciPy, PyQt, Scikit-learn, AdWords API
Tools
IPython Notebook, Sublime Text 2, IPython, StatsModels, MATLAB, Atom, Jupyter, GitLab
Paradigms
Business Intelligence (BI), Kanban, Functional Programming, Scrum
Storage
HDFS, Google Cloud, Redis, MongoDB, PostgreSQL, MySQL
Other
Data Engineering, Machine Learning, Data Mining, Deep Learning, Financial Engineering, Statistics, Data Visualization, Data Analysis, Bayesian Statistics
Platforms
Linux, Google App Engine, MacOS, Docker
Education
PhD Degree in Biomedical Image Processing
Masaryk University - Czech Republic
Master's (MBA) Degree in Business Administration
City University of Seattle - Seattle
Master's Degree in Statistics and Data Analysis
Masaryk University - Czech Republic