Kevin Drevitch

About Kevin Drevitch

With a strong foundation in statistics and data visualization, I excel in using Python to conduct A/B testing, hypothesis testing, and both descriptive and inferential statistics. My machine learning and data analysis expertise includes working with Scikit-Learn, NumPy, and Pandas to construct recommendation systems, build classification and regression models, and conduct exploratory data analyses. I am proficient in clustering methods, PCA, hyperparameter tuning, tree-based algorithms, and time series analysis. Additionally, I have experience in AI and deep learning, producing neural networks and natural language processing scripts. My technical toolkit also includes SQL, AWS, Azure, Docker, and various data analysis and visualization tools like Seaborn, Matplotlib, Power BI, and Tableau.

Experience

1. Data Scientist

   ColdSnap

   Oct 2023 — Present · Billerica, MA, US

   Collaborate directly with the President and VP leadership to identify and solve problems, optimize operations, and develop predictive models that enhance decision-making.• Build and maintain an ARIMA forecasting model to predict field consumption, enabling proactive resource planning.• Develop and implement machine learning models and statistical algorithms to confidently predict the likelihood of issues and classify their root causes, unsupervised learning to identify highest performing appliances, and a neural network to reduce customer waiting time by up to 15 seconds.• Engineer automated tools and dashboards to visualize key performance indicators, detect issues/anomalies, and provide actionable data-driven insights, including a tool that identifies and communicates product failures in the field within 5 minutes, improving customer success by enabling timely responses.• Drive scalable data pipelining and data storage initiatives, and develop optimal methods for collection and processing, ensuring quality, accuracy, and consistency across tools, analyses, and automated daily reports.