Tolulope Taiwo

I am a data scientist with hands-on experience delivering full-cycle machine learning projects, combining production work at VK with ongoing master's studies in Data Science. I focus on classical ML and NLP solutions that measurably improve product outcomes.

At VK, I developed models for gender prediction, age detection, and toxic comment categorization using datasets from hundreds of thousands of records, achieving tangible improvements such as a 15% lift in advertising targeting and automatic flagging of 85% of toxic content. My work spans feature engineering, model comparison (GBDT, Random Forest, CatBoost), hyperparameter tuning, and handling class imbalance.

I build with Python and a modern ML toolset—Pandas, Scikit-learn, LightGBM/CatBoost/XGBoost, HuggingFace, and deployment tools like Docker and FastAPI—while prioritizing reproducible pipelines and clear evaluation metrics. I am motivated by impactful NLP and classic ML problems and continuous professional growth.