Юрій

Yurii Polulikh AI Engineer
[открыть контакты](см. выше в блоке «контактная информация») [открыть контакты](см. выше в блоке «контактная информация»)

Lviv, Ukraine /in/yurii-polulikh /Yureeets

@yureeets /yuriipolulikh

I am a passionate AI Engineer with expertise Languages
in generative models, natural language
processing, and computer vision. I thrive in Ukrainian
designing deep learning architectures,
English
optimizing model performance, and solving
challenges across domains like image
recognition and language processing. Education
Committed to continuous learning, I aim to
Lviv Polytechnic National University
advance AI technologies and deliver
impactful results. 09/2022 – present | Lviv, Ukraine

Professional Experience
AI Engineer, Mod-y 11/2024 – present
Develop and improve the quality of generative models, including GANs and diffusion
models, while working on image recognition, natural language processing (NLP), and
dataset preparation.

Certificates
Machine Learning Specialization Introduction to TensorFlow for Artificial
Intelligence, Machine Learning, and Deep
Improving Deep Neural Networks:
Learning
Hyperparameter Tuning, Regularization
and Optimization Neural Networks and Deep Learning

Skills
Python TensorFlow PyTorch Analytical Skills
Active Learning Communication C++ SQL
NoSQL Computer Vision Data Processing Git
NLP Tableau Data Analysis Generative
Models

Yurii Polulikh 1/2
 Projects
Information Retrieval
Participated in Kaggle, dedicated to Ukrainian-language texts, aimed at developing natural
language processing (NLP) capabilities for this underrepresented language (Ukrainian).
Developed and implemented algorithms for efficient information retrieval in various datasets
of Ukrainian news, Wikipedia articles and social media posts.

Glasses Segmentation
Developed a deep learning model based on U-Net architecture that achieves 99% accuracy for
eyeglasses segmentation in images. Applied data preprocessing and augmentation techniques
to improve the model's performance. Demonstrated expertise in image segmentation using
convolutional neural networks (CNNs) to accurately predict masks. Achieved impressive
results on the task with 99% program accuracy on the validation set.

Tuberculosis Spread Simulation
Developed an agent-based model using Mesa to simulate TB spread based on real-world data.
Compared infection dynamics and intervention strategies across cities in Ukraine and
Europe, integrating regional data, dynamic health states, and visualizing outcomes.

Colorizing the images using GAN
Developed a Generative Adversarial Network (GAN) to colorize grayscale images. Designed
and trained the model on real-world datasets. Evaluated performance through visual and
quantitative metrics.

Yurii Polulikh 2/2