Data Science & Analytics Projects
Welcome to my portfolio repository, where I showcase my work across various programming languages and data analysis tools. This repository is organized into folders based on the tools and software I use regularly in my professional work.
Repository Structure
- Power BI: This folder contains projects and reports created in Power BI, showcasing my ability to develop insightful and interactive dashboards.
- Python: This folder contains Python scripts and Jupyter notebooks for data wrangling, analysis, and automation.
- Papers: This folder contains academic papers, research projects, and technical articles I have authored across various fields of study.
Each folder includes detailed documentation for the respective projects, providing context on the goals and methodologies used.
About Me
I am a Database Manager and Report Analyst currently working at Miami Dade College, where I oversee data analysis and reporting for the School of Science. I also serve as an Adjunct Faculty member at Miami Dade College, teaching classes in Intro to Artificial Intelligence and Machine Learning. I am responsible for designing and maintaining SQL databases, automating data processes with Python and Power Automate, and creating Power BI dashboards for decision-making. Additionally, I work with nonprofit organizations like the Belafonte Tacolcy Center to enhance data-driven strategies for client outreach and grant effectiveness.
Professional Skills:
- Data Analytics: SQL, Power BI, Tableau, Python
- Database Management: Azure SQL, AWS RDS (MySQL), Oracle
- Automation: Power Automate, Python Scripting
- Project Management: Leading data collection and reporting initiatives for academic departments and nonprofits
Featured Projects
- Power BI Dashboards: Interactive dashboards focusing on academic performance metrics, cohort analysis, and retention trends.
- Python Automation Scripts: Python projects designed to automate data processing, including scheduled tasks integrated with SQL databases.
- Database Development: Database solutions created in Azure SQL and AWS RDS, supporting student and nonprofit data tracking.
- Financial Planner App: Personal finance manager for income, expenses, and bank statement extraction with Google Gemini and Supabase integration.
Projects and Papers
This project was developed as part of the selection process for a Disney internship. It uses the Streamlit framework in Python to train a Logistic Regression machine learning model on historical data. The model predicts clients likely to abstain from purchases in the upcoming month, helping the marketing team target their campaigns and increase revenue.
Features:
- Logistic regression model to predict future purchases
- Upload CSV files to train the model and predict results
- Export processed data for reporting
2. Tutors Track
Tutors Track is a web application built with the Streamlit framework, hosted on Azure Web App, and connected to an Azure SQL Server. The application is designed for internal use, allowing students to track participants in tutoring sessions and manage attendance for classes or general events. It also includes an admin page for managing tutors, class codes, and generating reports.
Note: This is a private repository containing proprietary code developed for internal institutional use.
Features:
- Attendance tracking for tutoring sessions and events
- Admin dashboard to manage tutors, class codes, and generate reports
- Microservices architecture using Azure Functions to handle database requests, improving speed and reliability
- Hosted on Azure Web App and connected to Azure SQL Server for secure, cloud-based data management
- User authentication and role-based access control
- Automated reporting and data visualization
This paper offers a comparative analysis of two major international data privacy laws: the European Union’s GDPR and Brazil’s LGPD. It was developed as part of an Ethics in Cybersecurity course at Miami Dade College and demonstrates my ability to critically evaluate legal frameworks in cybersecurity and data protection.
Highlights:
- Comprehensive overview of GDPR and LGPD principles, rights, and enforcement mechanisms
- Detailed side-by-side comparison table
- Discussion of responsibilities for cybersecurity professionals regarding compliance and incident response
- Real-world enforcement cases illustrating practical applications
- Final reflections on global trends in data protection and privacy
This project explores the application of imitation learning to develop an AI agent capable of replicating expert gameplay strategies within a simulated ice hockey environment using SuperTuxKart.
Highlights:
- Developed multiple deep learning models (Model A, B, and C) to predict an expert agent’s behavior
- Focused on feature engineering, model training, and fine-tuning (learning rate adjustments, kart type analysis)
- Achieved significant performance improvements through model architecture refinement and strategic data preparation
- Demonstrated the effectiveness of deep learning for decision-making in dynamic simulated environments
This paper investigates the robustness of the ELECTRA-small transformer model in Natural Language Inference (NLI) tasks. Developed during my Natural Language Processing course, the work explores fine-tuning techniques and ensemble-based strategies to address dataset artifacts and improve model generalization.
Highlights:
- Fine-tuned ELECTRA on contrastive and adversarial datasets to expose model weaknesses
- Developed an ensemble-based training method for better handling of linguistic artifacts
- Detailed evaluation of model performance using accuracy and F1 scores
- Insights into improving robustness and reliability of NLI models
This article, written in Portuguese, was my undergraduate capstone project in Industrial Engineering. It analyzes the root causes of equipment failures at the Vila Velha Port Terminal and proposes a corrective action plan using quality management tools to increase asset availability and reduce unplanned downtime.
Highlights:
- Practical application of quality tools: Pareto chart, Cause-and-Effect diagram, Stratification
- Calculation of Key Performance Indicators such as MTBF (Mean Time Between Failures), MTTR (Mean Time To Repair), and Availability
- Identification of recurring failure modes in components like joystick and engine
- Action plan to enhance operational efficiency and reliability
Financial Planner is a personal finance web application built with Streamlit. It enables users to manage accounts, track income and expenses, upload PDF bank statements, and visualize their finances through an interactive dashboard. The app leverages Google’s Gemini Generative AI to extract transaction data from PDF statements and securely stores information in Supabase. User authentication is handled via Microsoft OAuth, ensuring data privacy and security.
Highlights:
- Add, edit, and delete accounts and transactions
- Upload PDF bank statements for automatic transaction extraction using AI Google Gemini
- Secure authentication via Microsoft OAuth (Azure AD)
- Real-time dashboard with financial summaries and charts
- Cloud data storage in Supabase for reliability and access
- Modular, component-based architecture for maintainability
- Docker support for easy deployment
- Step-by-step setup and test harness included in the repository
Certifications & Education
Along with my project work, I have completed various certifications and degrees that demonstrate my dedication to both business intelligence and data science. You can explore the details of these certificates and diplomas in the following categories:
- Degree: Includes diplomas such as my Bachelor’s in Industrial Engineering and Master’s in Science.
- Business Intelligence: Certifications related to BI tools and methodologies.
- Data Science: Certifications in Python, machine learning, and data science methodologies.
- Agile: Certifications in Agile project management and methodologies.
Each certificate includes details of the issuing institution, the skills covered, and validation information where applicable.
Resume
You can view my professional resume here:
👉 View My Resume