About Me

My name is Kelvin Santos Andrade

Data Analyst with +7 years of experience in retail and manufacturing, specialized in operational efficiency and business intelligence. Currently at Fiação Itabaiana, I transform data into strategies that drive revenue growth and cost reduction.

My Role

I develop and automate dashboards and reports integrated with Sankhya ERP, delivering actionable KPIs for Sales, Inventory, Shipping, and Finance. I support a Continuous Improvement culture (PDCA/Kanban) with reliable, real-time data.

Tech Stack

  • BI & Visualization: Power BI (DAX/Power Query), Advanced Excel
  • Data & Analysis: SQL (Oracle, PostgreSQL, SQLite), Python (Pandas, Scikit-Learn)
  • Methodologies: PDCA, Kanban, Data Governance, Kaizen

💡 My goal is to transform complexity into clarity, delivering insights that drive strategic decisions.

Skills

Programming & Databases

  • SQL for extraction, transformation, and analysis of relational data
  • Programming logic, data manipulation with Pandas, interactive visuals with Streamlit, ETL concepts, and cloud integration
  • Advanced Excel with Power Query and Pivot Tables

Statistics & Machine Learning

  • Dispersion measures (range, variance, standard deviation)
  • Data visualization with graphical tools
  • Outlier detection and treatment
  • Association analysis between variables
  • Descriptive, diagnostic, predictive, and prescriptive analytics
  • Machine Learning algorithms applied according to business models

Data Visualization

  • Building interactive dashboards in Power BI
  • ETL with Power Query and data modeling
  • Data storytelling and goal tracking with time intelligence functions
  • Pareto 80/20 analysis, cohort retention, cumulative metrics
  • Python visualizations with Matplotlib, Seaborn, and Plotly

Software Engineering

  • Version control with Git and GitHub
  • Project organization with clean code practices
  • Task automation and simple data pipelines
  • Project documentation using Jupyter Notebook
  • API development with Flask

Professional Experience

Where I worked and the impact I generated

Data Analyst - Fiação Itabaiana

August 2024 - Present · Ribeirópolis, Sergipe, Brazil

Creation, maintenance and automation of reports and dashboards integrated with ERP Sankhya, using SQL, Power BI (DAX/Power Query) and Excel to deliver actionable KPIs for Sales, Inventory, Shipping and Finance in a Continuous Improvement environment.

Key Responsibilities

  • Creation, maintenance and automation of reports and dashboards integrated with ERP Sankhya
  • Use of SQL, Power BI (DAX/Power Query) and advanced Excel to transform data into insights
  • Delivery of actionable KPIs for Sales, Inventory, Shipping and Finance
  • Support for Continuous Improvement routines (PDCA)
  • Focus on data quality, governance and storytelling with data

Skills

MySQL, Microsoft Excel, Power BI, SQL, DAX, Power Query, ERP Sankhya, Data Analysis, KPIs, Continuous Improvement

3+ Years as Administrative Manager

Main Activities

  • Monitoring KPIs for inventory, sales, and operational expenses
  • Building Power BI dashboards with storytelling and region/store segmentation
  • Extracting and cleaning data in SQL for insights and management reports
  • Automating spreadsheets and processes with advanced Excel and VBA
  • Supporting decision-making focused on expense reduction and EBITDA improvement
  • Managing reports on logistics, employee transport, and energy consumption

Results

  • Reduced operational expenses by 28% through logistics reorganization and fleet control
  • Saved R$ 10,000/month by optimizing refrigeration and energy usage
  • Improved inventory data accuracy by 15% using SQL and Excel
  • Implemented Power BI dashboards reducing managerial analysis time by 40%

1 Year as Receiving Supervisor

Main Activities

  • Invoice control and validation according to tax legislation
  • Team management and organization of goods receiving flow
  • Supervised operational processes and met deadlines
  • Created spreadsheets for product entry control and order checking
  • Trained and developed team members

Results

  • Reduced average receiving time by 25% by reorganizing operational flow
  • Improved fiscal accuracy by 18% using Excel cross-checking
  • Reduced operational errors by 40% with training routines

4+ Years as IT Assistant

Main Activities

  • Technical support for internal users with equipment, systems, and networks
  • Preventive and corrective maintenance of computers, printers, and POS
  • Monitored RUB system and resolved operational failures
  • Opened and managed technical support tickets
  • Updated software and systems per security policies
  • Supported network infrastructure, cabling, and access control

Results

  • Reduced recurring issues by 35% through standardized support processes
  • Decreased average ticket resolution time by 22% with scripts and checklists
  • Increased POS uptime to 98% with scheduled maintenance
  • Provided support to over 150 users and 26+ active POS units

Data Analysis Projects

Projects demonstrating technical skills and strategic vision

Warehouse Analysis Dashboard with Machine Learning

Warehouse Dashboard with Machine Learning

Intelligent system developed for predictive analysis and strategic management of industrial warehouses. Uses advanced Machine Learning techniques to predict demands, detect anomalies, and optimize maintenance processes. The dashboard offers an interactive interface with Streamlit, supporting Excel and CSV files.

  • Tools: Python, Streamlit, Scikit-learn, Prophet, Plotly, Pandas
  • ML Models: Linear Regression, Random Forest, Gradient Boosting, Prophet (Meta)
  • Features: 3-12 months forecasts, anomaly detection, criticality analysis, financial control
  • Highlight: NumPy 2.0+ patch, automatic data validation, chronologically ordered interactive charts
Heatmap Cohort SuperStore

Cohort Analysis – SuperStore

Using Excel, I developed a customer retention cohort analysis with real data from a national supermarket chain. Through data cleaning and structuring, I identified the best acquisition groups and behavior patterns, generating strategic insights for retention and sales.

  • Tools: Excel, VLOOKUP, Pivot Tables, Conditional Formatting
  • Result: May/2014 cohort had the highest retention after 6 months (11%)
  • Key Insight: Promotional campaigns positively impact repurchase behavior

RFM Analysis – Customer Segmentation

I classified customers based on their purchasing behavior using the RFM technique (Recency, Frequency, and Monetary). This analysis enables segmentation of the customer base to create personalized campaigns and maximize marketing ROI.

  • Tools: Excel (INDEX, MATCH, SUMIF, PERCENTILE)
  • Segments Created: Champions, Loyal, Promising, Lost, and more
  • Results: 40% of customers are potential loyal; only 1% in “Do Not Lose” group

DataLuz – SaaS Platform for E-commerce

I developed a fictional digital solution for small and medium-sized e-commerce businesses that struggle with data analysis. DataLuz offers intuitive dashboards, automated WhatsApp alerts, and customized reports. This project is part of the DS Community course and includes the full creation of a digital business model using AARRR strategies.

  • Tools: Product Strategy, AARRR Metrics, Customer Journey
  • Problem: Lack of time and technical knowledge for BI
  • Solution: API integration, automated reports, and simplified visualization
  • Highlight: Insights sent directly to the entrepreneur's WhatsApp

Loyalty Prediction with AI – Customer Classification

In this project, I developed a Machine Learning model using Decision Tree to predict the likelihood of customers signing up for loyalty cards (Aurora, Nova, and Star). The interface was built with Gradio, allowing anyone to simulate the classification of a new customer intuitively. Perfect for marketing and retention applications.

  • Tools: Python, Pandas, Scikit-Learn, Gradio, Matplotlib
  • Problem: Companies couldn't predict customer likelihood to join loyalty programs
  • Solution: Decision tree classification + interactive Gradio interface
  • Highlight: Interactive deployment with real-time tree visualization

Legal Data Simulation with Synthetic Data for Data Science

I created a realistic synthetic legal data generator, producing a complete dataset with 2,000 cases, parties involved, and procedural movements. This solution is ideal to showcase skills in data analysis, BI, and predictive modeling in a scenario that mirrors the real legal world.

  • Tools: Python, Pandas, Faker
  • Problem: Difficulty obtaining detailed legal data due to confidentiality
  • Solution: A script that generates a robust, realistic dataset in .csv format, ready for use in Power BI, Excel, or Python
  • Highlight: Ability to simulate a complex environment including claim values, risk provisions, legal stages, and responsible lawyers

Accommodation and Pricing Analysis – Airbnb NYC

A full Data Analysis project using a large Airbnb dataset from New York City. The goal was to turn thousands of records into strategic insights about accommodations, pricing, and location patterns through a dynamic and visually intuitive Power BI dashboard.

  • Tools: Power BI, Power Query, DAX
  • Highlights: Pareto chart (Top 100 neighborhoods), geographic map, time analysis, and performance indicators (KPIs)
  • Key Insight: Identification of neighborhoods with the highest concentration of premium listings and seasonal price patterns

📝 Blog

Insights on Data Analysis and Machine Learning

I share practical cases, trends, and learnings about data analysis, machine learning, and technology in industry. Check out the latest article:

📊 Warehouse Dashboard with Machine Learning: A Success Case

October 17, 2025 | 15 min read

How I transformed an inventory management problem into an intelligent solution using Power BI, Python, and Machine Learning. Discover the challenges, technical solutions, and measurable results of this project in the textile industry.

Main topics: Demand Forecasting, Anomaly Detection, Inventory Optimization, 320% ROI

🎯 Case Studies

Real projects with measurable results

🔬 Machine Learning

Practical ML applications

🏭 Textile Industry

Trends and technology

Contact

Feel free to get in touch

© 2025 Kelvin Andrade. All rights reserved. Data Analysis Portfolio developed by Kelvin Andrade. Design based on a template by HTML5 UP. Images from Unsplash.