Артем

Artem Khomytskyi
[відкрити контакти](див. вище в блоці «контактна інформація») | linkedin | https://github.com/ArtemKhomytskyi

Skills
Languages & Tools: Python, SQL, R, HTML (Bootstrap), Assembly
Machine Learning: Supervised / Unsupervised models, Deep Learning (CNN, Transfer Learning), Time Series (ARIMA, ETS)
Data & Systems: MongoDB, Docker, Databricks, Power BI
Other: Git, n8n automation, AI assistants, Discord API
Languages: English (B2), Ukrainian (native), Russian (native), Polish (B1), Portuguese (A2)
Experience
Community Manager, Liquorice – Lisbon, PT February 2024 – Present
● Designed and launched a crypto-native Discord ecosystem with optimized IA, onboarding flows, and safety systems,
improving user retention and feedback integration.
● Built AI-driven onboarding and FAQ assistants via n8n, reducing average response time and ensuring 24/7 community
engagement.
● Introduced Discord leveling and gamification features, resulting in higher weekly message volume and sustained
community growth.
Executive Assistant, Velvet Formula – Lisbon, PT Jul 2023 – Feb 2024
● Rebuilt the company website with HTML (Bootstrap), improving visual consistency and reducing bounce rates by 15%.
● Integrated generative AI tools into content and design workflows, increasing article output from 1 to 3 per week.
● Managed investor data and designed automated Excel tracking formulas during early funding rounds, helping secure
additional project financing.
Projects
Time Series Forecasting – Apple Stock (R) Github
● Implemented a full Box–Jenkins pipeline (ARIMA) for Apple stock (2015–2024), including preprocessing, stationarity
checks, AICc model selection, and residual diagnostics.
● Benchmarked ARIMA vs. ETS and STL+ETS models, improving forecast accuracy through rolling-origin evaluation.
Big Data Analysis (PySpark, Databricks) Github
● Processed large datasets in Databricks using PySpark; applied regression and random forest models to analyze crime,
wine quality, and marketing data.
Deep Learning – Skin Cancer Classification (CNN) Github
● Built CNN models and applied transfer learning on the HAM10000 dataset to classify skin lesions; optimized
preprocessing and achieved strong accuracy metrics.
Relational Database Design (SQL) Github
● Designed and implemented a relational database for the 2027 Women’s Football World Cup; optimized SQL queries
for efficient data retrieval and integrity.
Customer Segmentation & Targeted Promotions (ML) Github
● Conducted customer segmentation with PCA and KMeans, enabling data-driven marketing campaigns for retail
clients.
Education
Nova University of Lisbon – Bachelor in Data Science Expected May 2026
Primorsky Lyceum – Physics & Applied Mathematics May 2022