Yevhen

Євген Томаши
Junior Data Scientist
Data scientist with 1 year of experience, showcasing expertise in Python, machine learning, computer
vision and NLP. Accomplished in steering data pipelines, deploying models, and effectively translating
technical insights for impactful business outcomes.

[відкрити контакти](див. вище в блоці «контактна інформація») [відкрити контакти](див. вище в блоці «контактна інформація»)

Lviv, Ukraine [відкрити контакти](див. вище в блоці «контактна інформація»)

github.com/batiamida

WORK EXPERIENCE SKILLS
Junior Data Scientist at Python SQL (PostgreSQL, MySQL)
Empowero sp
1 year Deep learning (Computer vision, NLP, Generative AI
Achievements/Tasks basics)
Contributed to a ground-up project at a small product
Matplotlib, seaborn, Plotly TensorFlow, PyTorch
company, emphasizing effective time management and
collaborative teamwork.
OpenCV Scikit-Learn, Statsmodels
My tech stack in the company : python, pandas, numpy,
SciPy, jupyter-notebook, matplotlib, seaborn, plotly, sklearn, AWS ( EC2, S3, Lambda, SageMaker, CloudWatch )
statsmodels, tensorflow, PostgreSQL, SQLAlchemy, alembic.
The most used models in the project : linear regression, ORM and migrations (SQLAlchemy + alembic)
logistic regression, knn, tree-based.
Docker
Experience : Worked with big data, handling datasets with
over a million rows (built data pipelines); Conducted research
to find data that addresses business problems; Used tree-
based algorithms to define rules that would help business;
Optimized database queries. PERSONAL PROJECTS
COVID mask detector (2023)
Worked with OpenCV during pandemic to build mask detector
EDUCATION system, to detect mask on people in real time.
Used keras to train my custom classifier with DNN, Convolutional
and pooling layers.
BACHELOR degree (3d grade)
Used transfer learning (pre-trained model with
Lviv Polytechnic National University cv.CascadeClassifier) to detect faces, this step needed some
2021 - 2025, understanding of how to preprocess image in order to detect
faces even with mask on and which model to use.
SoftServe AI "propaganda detector bot" project with mentor
Teamwork in university with Utilized F1 score, precision,
mentor from SoftServe and recall for evaluating
helped me to the slightly unbalanced
data.
CERTIFICATES
Developed a Telegram bot Collected and processed AWS Cloud Technical Essentials
to classify messages as data and implemented a
propaganda or not. logistic regression model Analyze Datasets and Train ML Models using AutoML
due to deadlines didn't
have enough time to Object Localization with TensorFlow
explore further deep
learning models. Visualizing Filters of a CNN using TensorFlow

Data Wrangling, Analysis and AB Testing with SQL

LANGUAGES
Ukrainian English
Fluent B2

Chinese Japanese
Beginner Beginner