Ilya Ezepov, Developer in Moscow, Russia
Ilya is available for hire
Hire Ilya

Ilya Ezepov

Verified Expert  in Engineering

Deep Learning Developer

Moscow, Russia
Toptal Member Since
May 29, 2019

Ilya is a data-obsessed analyst with experience in developing large-scale machine learning solutions and managing tech teams. He has proficient knowledge in statistics, mathematics, and machine learning combined with the scientific background. Ilya also loves acquiring new knowledge and sharing it with others.


Jupyter, LightGBM, PyTorch, Python, Scala, Artificial Intelligence (AI)...
TensorFlow, Keras, MapReduce, Jupyter, Python, Artificial Intelligence (AI)...




Preferred Environment

Git, Vim Text Editor, Jupyter, PyCharm, MacOS, Linux

The most amazing...

...ML system I've built is an unsupervised word2vec-like tool that allowed to process petabytes of data from various sources and merge it into the semantic space.

Work Experience

Senior Data Scientist

2017 - 2019
  • Integrated machine learning (ML) into the main pricing engine; implemented demand optimization in order to maximize key-performance indicators (KPIs).
  • Taught internal ML courses and organized ML competitions.
  • Translated business KPIs to ML projects and explaining data insights to business teams.
  • Working on the core-price algorithm and on the various smaller ML-related projects.
  • Developing ML end-to-end pipelines; starting from data logging and processing to model training to deployment to model monitoring.
Technologies: Jupyter, LightGBM, PyTorch, Python, Scala, Artificial Intelligence (AI), Machine Learning, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Data Pipelines

Data Scientist | Team Lead

2015 - 2017
  • Developed a user profile for targeted advertising; determined a user's social profile (gender, age, income) and interests with machine learning based on internet behavior.
  • Built a large-scale unsupervised look-alike system, for a fast automated search for similar users (US patent pending 2019/0034535).
  • Led a team of three data scientists and three developers in the research, development, and integration of solutions.
  • Worked on an unsupervised user segmentation as a part of Yandex Audiences project.
  • Implemented large-scale data processing and machine learning with Python as a wrapper for a proprietary MapReduce engine.
Technologies: TensorFlow, Keras, MapReduce, Jupyter, Python, Artificial Intelligence (AI), Machine Learning, Data Pipelines


2014 - 2017
Center of Mathematical Finance
  • Taught a non-profit course on machine learning and data analysis to about 100 participants that focused more on the financial data analysis.
  • Covered various topics in ML: from the introduction to numerical optimization to recent advances in neural nets.
Technologies: Python


2014 - 2015
Moscow Exchange
  • Productionized the data import pipeline from various financial sources to generate the weekly market review for the leadership.
  • Supported a legacy SQL and worked on improving it which resulted in ~two times speedup in processing time.
  • Standardized the process of fetching the market news and generating reports from them.
Technologies: Excel VBA, SQL, Data Analysis, Data Visualization

Zebra Classroom
Zebar Classroom is an edtech side project for real-time classroom feedback and management. I used React for the front end and Google Firebase on the back end.

Kaggle Competition Master
I compete in coding competitions and am a Kernel expert. The highest place I've achieved so far was 902nd on the global competition leader board. is a side project for intermediate/advanced English learners. It allows a user to learn words before watching a movie. The front end is built in React (TypeScript) and back end is in Python (SpaCy as the main NLP engine).
2014 - 2016

Master's Degree in Mathematical Methods in Finance

Lomonosov Moscow State University - Moscow, Russia

2010 - 2014

Bachelor's Degree in Material Science

Lomonosov Moscow State University - Moscow, Russia


PyTorch, Scikit-learn, Pandas, NumPy, Keras, TensorFlow, XGBoost, React


Jupyter, Git, PyCharm, Vim Text Editor


Python 3, Python, SQL, Excel VBA, Scala, Java, JavaScript, TypeScript


Jupyter Notebook, Linux, MacOS


Data Pipelines, MongoDB




Functional Programming, MapReduce, Test-driven Development (TDD)


Neural Networks, Deep Learning, Convolutional Neural Networks (CNN), Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Word2Vec, Large Language Models (LLMs), Data Visualization, Data Analysis, Materials Science

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.


Share your needs

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

Choose your talent

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

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