Josef Ondrej, Developer in Prague, Czech Republic
Josef is available for hire
Hire Josef

Josef Ondrej

Verified Expert  in Engineering

Mathematical Modeling Developer

Prague, Czech Republic

Toptal member since November 12, 2019

Bio

Josef has a background in mathematics with a specialization in statistics and probability theory. He has five years of experience with applying his skillset to modeling various real-world problems ranging from in-play odds calculations for sports matches to building AI solutions in the field of conversational agents.

Portfolio

IBM Watson
Jenkins, Kubernetes, Docker, Java, Python
Tipsport
Theano, Keras, TensorFlow, Pandas, NumPy, R, Selenium, C#, Java, Python

Experience

  • Mathematical Modeling - 9 years
  • Statistics - 7 years
  • Python - 5 years
  • NumPy - 5 years
  • Deep Learning - 3 years

Availability

Part-time

Preferred Environment

GitHub, PyCharm, Ubuntu, Linux

The most amazing...

...thing I've worked on was a model for calculating in-play odds for tennis and volleyball matches in a major gambling company.

Work Experience

Research Scientist

2018 - PRESENT
IBM Watson
  • Contributed to research on how dialog agents can be designed using a declarative paradigm and planning technologies like PDDL.
  • Implemented a production-grade system that enables dialog agents to automatically improve based on their interactions with the end-users.
Technologies: Jenkins, Kubernetes, Docker, Java, Python

Mathematician

2014 - 2018
Tipsport
  • Developed mathematical models for calculating in-play odds for various sports matches including tennis, (beach-)volleyball, badminton and horse racing.
  • Scrapped website to collect data for subsequent analysis and model parameters estimation.
  • Developed a model for predicting the risk rating of clients based on their previous betting activity.
  • Worked on an image recognition model for tracking the position of a ball in table tennis.
Technologies: Theano, Keras, TensorFlow, Pandas, NumPy, R, Selenium, C#, Java, Python

Experience

Live Odds Models

If you want to bet on (beach) volleyball, badminton, or tennis doubles at www.tipsport.cz, you are likely betting on odds calculated using a mathematical model that was developed by myself.

Plotting Text in Spectral Analyzer

https://github.com/josefondrej/word2wave
This was a toy project that immediately popped into my mind when I first installed a sound spectral analyzer on my phone.

MNIST explained

https://github.com/josefondrej/mnist
This one was created as a learning source for a friend who wanted to explain the basics of deep learning.

Education

2013 - 2015

Master's Degree in Statistics, Probability Theory and Econometrics

Charles University - Prague, Czech Republic

2010 - 2013

Bachelor's Degree in General Mathematics

Charles University - Prague, Czech Republic

Certifications

FEBRUARY 2018 - PRESENT

Machine Learning

Coursera

FEBRUARY 2018 - PRESENT

Sequence Models

Coursera

NOVEMBER 2017 - PRESENT

Convolutional Neural Networks

Coursera

OCTOBER 2017 - PRESENT

Structuring Machine Learning Projects

Coursera

OCTOBER 2017 - PRESENT

Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization

Coursera

OCTOBER 2017 - PRESENT

Neural Networks and Deep Learning

Coursera

Skills

Libraries/APIs

NumPy, Theano, TensorFlow, Keras, Pandas, Matplotlib

Tools

GitHub, PyCharm, IntelliJ IDEA, Jenkins, Hidden Markov Model, ARIMA

Languages

Python, Java, C#, R, SQL, CSS, HTML, Bash

Frameworks

Flask, Selenium

Paradigms

Declarative Programming, Imperative Programming, Agile Software Development, Concurrent Programming

Platforms

Linux, Ubuntu, Docker, Kubernetes

Storage

SQLite, PostgreSQL, MongoDB, JSON

Other

Bayesian Statistics, Bayesian Inference & Modeling, Statistics, Probability Theory, Regression Modeling, Robust Regression, Logistic Regression, Polynomial Regression, Regression, Mathematical Modeling, Markov Chain Monte Carlo (MCMC) Algorithms, Deep Learning, Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Machine Learning

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