Skip to content

Hello, I'm Renan Toyoyama 👋

Data Engineer · Data Scientist · Data Analyst

About me

Data Engineer and Data Scientist with end-to-end experience — from ETL/ELT pipelines to predictive modeling — working in Big Data environments with Spark, Hadoop, Hive and Databricks. Designed and delivered critical data solutions in production at the Secretariat of Finance of Pará (SEFA), including dbt layered architectures for fiscal risk analysis and internal Python packages for corporate system integration.

I combine technical expertise in Python, Airflow, DBT, Oracle and CI/CD with analytical vision to transform data into real business value.

🏢 Professional Experience

Data Engineer / Data Scientist Secretariat of Finance of the State of Pará

  • Built fiscal risk analysis pipelines to identify irregular taxpayers based on corporate ties and registration history
  • Developed internal Python packages with Poetry, uv and Docker, published via GitLab CI/CD
  • Orchestrated pipelines with Apache Airflow and Apache Oozie, with failure monitoring and controlled reprocessing
  • Built Machine Learning models (supervised and unsupervised) with Scikit-learn for fiscal risk classification
  • Worked in agile methodology (Scrum/Kanban) with Jira task management

Process Analyst

  • Identified bottlenecks in inventory, production and logistics
  • Created and standardized operational procedures
  • Monitored KPIs: delivery time, productivity and inventory control
  • Implemented process automation systems

Network Technician

  • Installation and maintenance of fiber optic networks
  • Configuration of routers, ONUs and network equipment

Junior Systems Analyst

  • Developed the automated PROEXT system (Sisproxt)
  • Trained public servants on the tool

Intern — Information Systems Tutorial Program

  • Academic tutoring in Information Systems courses
  • Participated in system development, websites and scientific articles
  • Presented at academic and professional conferences

  • Currency ETL Pipeline

    Automated ETL pipeline with Airflow, PostgreSQL and Docker. Collects USD, EUR and BTC exchange rates in real time.

    Python Airflow PostgreSQL Docker

  • Credit Score — Naive Bayes

    Predictive credit risk classification model with 98% accuracy and excellent generalization capability.

    Python Scikit-learn Pandas

  • Decision Tree Classification

    Supervised classification with Gini criterion, feature importance analysis and comparison with Naive Bayes.

    Python Scikit-learn Matplotlib

  • Real Estate Regression

    Rent price prediction with linear regression, outlier treatment and overfitting analysis.

    Python Scikit-learn Statsmodels

  • Customer Segmentation — K-Means

    Unsupervised clustering of 200 customers into 5 segments (silhouette 0.555) with marketing strategies for each group.

    Python Scikit-learn Pandas Docker

  • Titanic — Final Model Duel

    Hard voting ensemble with leak-free pipelines and a group survival feature. CV 0.8503 and 0.77751 on Kaggle, with 65 tests and Docker.

    Python Scikit-learn XGBoost Docker

  • Supermarket Sales Analysis

    Full EDA on supermarket sales data — consumption patterns, seasonality and strategic insights.

    Python Pandas Seaborn

  • E-commerce Dashboard

    SQL pipeline with INNER JOIN + self-contained HTML dashboard. 7 interactive visualizations with filters by gender, state and category.

    Python Pandas SQLite Chart.js

  • EduConnect SGE

    Educational Management System in Java with 3-layer architecture and SOLID principles.

    Java OOP SOLID

🛠️ Main Stack

Apache Airflow · Apache Spark · DBT · Hadoop / HDFS · Hive · HBase · Impala · Oozie · Docker · GitLab CI/CD

Scikit-learn · TensorFlow · Pandas · NumPy · Statsmodels · Matplotlib · Seaborn

PostgreSQL · Oracle Database · SQLAlchemy · Databricks · Impala

Python 3.13 · SQL · Shell Script · Java · Git · uv · Poetry · Ruff · Pytest

📫 Contact

Currently working as Data Engineer at the Secretariat of Finance of Pará (SEFA). Open to conversations about new challenges in the data field.

LinkedIn GitHub E-mail