Paula Obeng

I am a data analyst with a BSc. in Computer Science who builds predictive models and interactive analytics platforms. I focus on turning complex data into actionable insights that support clinical and business decisions.

I have engineered a cardiovascular disease prediction model using CNNs with 94% accuracy and deployed a motorcycle safety detection model (YOLO + CNNs) achieving 92% helmet detection accuracy. I also optimized an XGBoost classifier to 0.89 ROC AUC for term deposit prediction while using SMOTE to reduce false negatives.

I design and deploy dashboards and analytics platforms that scale — including a global heart disease analytics platform processing 50K+ records from 194 countries, Power BI dashboards for e-commerce and freight analyses, and Plotly-Dash real-time safety KPIs. My work combines ETL workflows, forecasting tools, choropleth visualizations, and KPI development to inform strategy and policy.

I volunteer on safety and training initiatives, collaborating with field teams to translate data into advocacy and operational change. I seek to contribute to mission-driven teams where predictive analytics, clear dashboarding, and continuous model improvement drive measurable impact.