Vladyslav Hamolia

State Scientific and Research Institute of Cybersecurity Technologies and Information Protection

Applied AI/ML projects in the defense technology domain, with most details subject to NDA. Work focuses on computer vision, structured data analysis, trajectory prediction, and audio recognition for military-related use cases.

• Computer Vision: detection and recognition of aerial objects, including UAVs and FPV drones.
• Trajectory Prediction: analysis of structured sensor data to predict aerial object movement.
• Audio Recognition: classification of military-related sound sources.
• Synthetic Data Generation: creation of synthetic datasets to support model training and evaluation.

MacPaw

Advanced AI/ML solutions for products used by millions of iOS and macOS users: defining ML strategy, leading applied R&D, and developing production-ready machine learning systems across NLP, computer vision, predictive modeling, and privacy-preserving AI.

• Computer Vision: semantic segmentation, object detection on photos and videos, image classification, and image reconstruction/enhancement (image inpainting).
• NLP: text classification, metric learning, information extraction and processing, Multi-Agent systems design, RAG pipeline development, and LLM fine-tuning.
• Predictive Modeling: forecasting customer lifetime value (LTV) and customer churn.
• Federated Learning: training neural networks for computer vision directly on user devices, improving models while preserving data privacy.

Kyiv School of Economics

Created and launched a Bachelor's program in Artificial Intelligence aligned with modern industry standards, emerging AI trends, and labor market needs. Designed a curriculum combining theoretical foundations with practical AI engineering skills. The program currently has 140 students. I also support collaboration between MacPaw and KSE to strengthen industry-academia connections in the technology sector.

FairmarkIt

Led development of a comprehensive search solution for platform efficiency in the Supply Chain sector, heading a team of Data Scientists and ML Engineers. Developed a multilingual semantic search algorithm to optimize search across various languages.

• Deep learning-based search engine with Learning-To-Rank functionality.
• Domain-specific hierarchical text classification with custom BERT-based models.
• ML infrastructure on AWS; scalability and reliability through best-practice adoption.

Eleks

Designed and developed data science solutions for a diverse range of clients across multiple industries.

• Embedded Systems: optimized neural networks for resource-constrained devices via quantization, pruning, and inference acceleration; built an internal CV framework similar to ONNX.
• Security: robust object tracking for video surveillance using SORT, DeepSORT, and YOLO-based detectors.
• Healthcare: semantic segmentation for medical image analysis with U-Net and FCN architectures, supporting clinical decision-making.

SoftServe

Developed and enhanced a large-scale American e-commerce project — from software design to integrating machine learning technologies.

• Robust web-scraping systems with asynchronous data retrieval and optimized data pipelines.
• Python best practices: modular design, unit testing, continuous integration.
• NLP models automating text data processing, integrated into the existing project ecosystem.