Python projects
Pet Market Insights
Problem
Transforming raw, unstructured records about an animal population into understandable knowledge. The central challenge was applying Exploratory Data Analysis (EDA) techniques to extract valuable insights into physical and demographic characteristics, enabling pet industry companies to make decisions based on real market and health trends.
Solution
Development of an analysis pipeline using the Python language for data structuring, cleaning, and visualization:
-
Manipulation and Cleaning (Pandas): Loading CSV files, integrity verification, and handling of null data;
-
Structured Aggregation: Grouping information by categories (race and gender) to enable statistical calculations;
-
Advanced Visualization: Implementation of bivariate visualization techniques to understand correlations between variables.
Result
The analysis yielded fundamental findings through three main techniques:
-
Age Distribution: Use of histograms to identify the predominant age range, allowing for the segmentation of health campaigns;
-
Weight Metrics by Breed: Bar charts that facilitate the immediate identification of breeds with a tendency towards overweight, assisting in nutritional standardization;
-
Coat Color and Gender Correlation: Count charts that reveal adoption preferences or market trends based on animal aesthetics;
-
Market Intelligence: Cross-referencing data to target the marketing of specific products to identified subgroups.
Tool used
-
Python: Programming language for data science;
-
Pandas: Library for manipulating and processing dataframes;
-
Matplotlib & Seaborn: Libraries used to create graphs with high aesthetic rigor and visual clarity.
Data source
-
Public Dataset (Kaggle): Database containing detailed records about cats (breed, age, weight, color, and gender).
Sales & Profit
Problem
Lack of clarity regarding sales volume and the real financial impact per product category in a retail operation. The challenge was to transform a raw list of transactions into a strategic analysis that would allow identifying which items have the highest inventory turnover and which categories are the pillars of the company's revenue.
Solution
Development of a Python solution focused on data modeling and automated commercial performance calculations:
-
Data Structuring (Pandas): Creation of DataFrames to organize and clean the raw structure of transactions;
-
Automatic Calculations: Generation of calculated revenue columns (Quantity × Unit Value) to enable the analysis of financial weight by category;
-
Technical Visualization: Use of a coordinated color palette to facilitate the reading of complex KPIs, focusing on professional aesthetics.
Result
The analysis provided fundamental insights into the financial health of the business:
-
Inventory Turnover Identification: It was found that the Accessories category (Keyboards/Mice) has the highest sales volume, being essential for daily cash flow;
-
Revenue Dominance: It revealed that the Electronics category dominates total revenue due to its high unit value and margin, despite its lower sales volume;
-
Upsell Strategy: Implementation of recommendations to offer accessories at the time of electronics purchase, optimizing the average ticket price;
-
Inventory Management: Definition of priorities for frequent replenishment of high-turnover items, avoiding stockouts.
Tool used
-
Python: Main technology for the overall development of the project;
-
Pandas: Library used for data manipulation, cleaning, and aggregation;
-
Matplotlib & Seaborn: For generating graphs with professional aesthetics and coordinated palettes.
Data source
-
Operational Database: Dataset simulating the sales flow of a retail operation, containing records of products, quantities, unit values, and categories.
