Gerardos_Portfolio

Hello there! 👋 I am Gerardo

📊 Data Analyst | Problem Solver | Storyteller with Data | SQL Queries | Python (scripts, libraries) | Pandas | Numpy | Seaborn | Matplotlib | Pivot tables | Dashboards

Data analyst with strong skills in Excel, Google Sheets, SQL, and Python, with prior experience in administrative areas. Passionate about solving problems using data. I have worked with Excel, Google Sheets, and SQL, focusing on transforming data into clear and actionable insights. I am currently strengthening my Python skills with libraries such as Pandas, NumPy, Matplotlib, Seaborn, as well as Power BI and Tableau, in order to move into data analysis–focused roles.

• Programming languages: Python, SQL (Oracle), Unix
• Libraries/Frameworks: Pandas, NumPy, Matplotlib, Seaborn, FastAPI, Django, OOP
• Analysis and Visualization: Dashboards, Pivot Tables, A/B Testing
• Tools: Microsoft Office, Google Sheets, Docker
• IDE: PyCharm, Visual Studio Code, Cursor
• Management: Jira, Slack
• Languages: Spanish: Native, English: Proficient

Projects

ConnectaTel – Customer Usage Analysis | Triple Ten 2026

This project focuses on understanding how customers actually use mobile services (calls and messages) in order to detect usage patterns, identify atypical behaviors, and generate actionable customer segments for a telecommunications company.

• Objective: To analyze mobile service usage in order to identify behavioral patterns, detect anomalous activity, and segment customers based on their communication habits to support better commercial decision-making.

• Data: Three integrated datasets containing mobile plans, user profiles (age, city, registration date, plan), and detailed activity logs for calls and SMS. Data preparation included handling missing values, correcting invalid entries (e.g., age = -999), imputing missing locations, and validating date consistency.

• Techniques: Data cleaning, exploratory data analysis, and customer segmentation using Python (Pandas, NumPy, Matplotlib, Seaborn). Outlier detection using boxplots and the IQR method, descriptive statistics, and behavioral segmentation based on usage intensity (calls and messages).

• Result: The analysis identified distinct customer usage segments and highlighted behavioral differences across plans and age groups. Outlier patterns in messaging and usage intensity suggested potential atypical behavior or data anomalies, providing insights for optimizing mobile plan design and improving ConnectaTel’s targeting strategy.

Urban Mobility and Economic Productivity in LATAM Cities | Triple Ten 2026

To identify which cities to invest in transportation infrastructure in order to improve productivity and population well-being.

• Objective: To identify which cities to invest in transportation infrastructure in order to improve productivity and population well-being.
• Data: A unique, curated dataset was built from two different sources, applying data cleaning, standardization, and validation processes to heterogeneous variables.
• Techniques: Data manipulation and analysis using Python (Pandas, NumPy, Matplotlib, Seaborn), SQL queries, and descriptive statistics; process documentation in Jupyter Notebook with visualizations and an executive report.
• Result: It was demonstrated that there is no simple linear correlation between a city’s GDP per capita and its level of vehicular traffic, which made it possible to rethink investment criteria toward variables more relevant for public decision-making.

E-commerce Funnel Retention Analysis With SQL | Triple Ten 2026

This project focuses on analyzing the full user funnel and retention performance using SQL.

• Objective: To optimize the user lifecycle by analyzing the conversion funnel and retention rates to identify friction points and revenue leakage within the platform.
• Data: Processing and cleaning of relational e-commerce datasets (funnel and retention), integrating millions of records of navigation events, transactions, and signup dates segmented by country and device.
• Techniques: Advanced SQL (extensive use of CTEs, complex JOINs, and aggregation functions) for multi-stage Funnel Analysis and Cohort Analysis; calculation of conversion KPIs and retention metrics (D7, D14, D28).
• Result: Identified a critical technical failure in the checkout process for 3 countries (0% conversion) and an 85% drop-off in purchase intent (“Add to Cart”), driving the prioritization of urgent technical fixes and a targeted CRM reactivation strategy to address user churn at Day 28.