Nikolay Kozlovskiy

• Developing advanced Spark components for synthetic data generation to evaluate methodologies and software efficiency.

• Designed and led the development of an automated pipeline for internet search and analysis, extracting article texts from Google News and conducting content assessments using large language models, including ChatGPT and Llama • Assisted in EUDR compliance by processing data in Google Cloud Platform and R, and trained the TransUNet model on AWS infrastructure for forest classification •Skills: Natural Language Processing, Computer Vision, PyTorch, TensorFlow, OpenAI, Ollama, Google Cloud, AWS, R, Docker

• Developed machine learning models for predicting Annual Average Daily Traffic on Estonian roads, simplifying data science concepts for non-technical stakeholders and presenting demo solutions. • Led a deep learning project using Graph Neural Networks for traffic prediction, overseeing data preprocessing, model training, deployment, and monitoring. • Conducted socio-demographic data analysis with SQL, implementing anomaly detection, correlation, and time-series analyses.

• Gained hands-on experience in engaging with businesses and customers, leading teams, developing research strategies, collecting data, and extracting insights. Skilled in writing reports and presenting results to clients and at conferences and forums. • Demonstrated expertise in a variety of methodologies, including surveys, focus groups, and content analysis of documents and online reviews, with a strong emphasis on data analysis and visualization using Microsoft Excel, SPSS, and Power BI.

Courses: Machine Learning, Neural Networks, Transformers, Data Science for Urban Mobility, Databases, Business Data Analytics, Geospatial Analysis with Python and R