Arthur Victor
As a data science and analytics enthusiast, I possess a robust understanding of data-driven concepts and business intelligence. Leveraging a versatile toolkit that includes Power BI, SQL, Excel, and Python, I extract valuable insights and steer data-driven initiatives. My expertise spans data analysis, machine learning, dashboard development, report generation, and various other data-related tasks. This proficiency enables me to deliver impactful results, unlocking the full potential of data-driven strategies.
Selected projects in data science and machine learning
A complete churn analysis
The problem of customer churn represents a central impasse for any telecommunications company. It is less costly to retain current customers than to acquire new ones.
I tackled this issue using Python as the primary tool.
Key Highlights
- Data Exploration (EDA): Conducted a thorough EDA, revealing vital insights:
- Customers with annual or biennial contracts tend to be more loyal.
- Monthly payment and manual payment methods are linked to higher churn rates.
- Model Development: Implemented a logistic regression algorithm after extensive data preprocessing and engineering.
- The model successfully identified 72% of potential churn cases during validation.
- Business Impact: Calculated the expected value of $95 per customer by applying the model to real-world scenarios.
- Provides actionable insights for informed decisions on customer retention strategies.
Customer Segmentation
Have you ever wondered how grocery stores determine which products to place where or how to create targeted marketing campaigns? The answer lies in customer segmentation!
I applied clustering techniques to a dataset of grocery store customers to group them into distinct segments based on their purchasing behavior. Through data analysis, I was able to describe the unique profile of each cluster, including their demographic information, preferred products, and shopping habits. This information can be used by grocery stores to optimize their inventory management, personalize their marketing efforts, and ultimately enhance the customer experience.
Brazilian Ecommerce Olist
In the dynamic landscape of e-commerce, the Brazilian E-Commerce Olist Analysis project stands as a guiding light, illuminating the path to data-driven decision-making. Just like skilled artisans, we meticulously explore the treasure trove of data provided by Olist, unearthing valuable insights that extend beyond mere sales figures.
Equipped with SQL and Power BI, our journey takes us deep into the heart of e-commerce intricacies, revealing logistics efficiency, customer satisfaction, and sales trends.
Skills-based projects
- Deployment - Flash/Gradio/Ngrok
- A Web-Scraping application
- Power BI dashboard repository
- Extracting Bovespa Stocks Data
- Email Report Automation
- Sentimental Analysis NLP
- Song-Popularity-Prediction