Data Scientist · toward ML Engineering & MLOps
Agustín Formenti
I analyze data and build end-to-end Machine Learning and AI systems — from idea to production.
Always building. Always learning.
Who I Am
About Me
Data Scientist with an engineering background and a solid math foundation, backed by real industry experience. I combine data analysis, applied statistics, model building, and process automation to solve real problems — with a clear direction toward ML Engineering and MLOps.
Engineering-rooted training — same rigor, now applied to data and models.
- Python
- SQL
- Machine Learning
- scikit-learn
- XGBoost
- PyTorch
- pandas
- R
- Docker
- SAP
- Power BI
- n8n
- Git
- TypeScript
- React
- PostgreSQL
Background
Experience
2026 — Present
Data Analytics & Internal Audit (Internship) · ArcelorMittal Acindar
Data analysis, internal audit, and process automation in a large-scale industrial environment, within the Global Assurance team.
- End-to-end design and development of an internal web system for Data Analytics Working Papers documentation, used by the Global Assurance team — Flask, SQLAlchemy, Bootstrap 5, SQLite; multi-role approval workflow, PDF/Excel export, user management, and automatic backup.
- Automation with Python and generative AI: text extraction from images with automatic upload to Excel, cutting down manual processing.
- Finished-goods stock aging analysis with Python and Power BI, and identification of production planning inefficiencies.
- Data extraction and management via SAP; development of automated reports for tracking operational KPIs.
- Python
- SQL
- Power BI
- SAP
- Flask
- Generative AI
- SOX / Internal Audit
2024 — Present
B.S. in Data Science · UCA Rosario
In progress, GPA 8.13/10. Training in Data Science, Machine Learning, and statistics — the foundation for a career in ML Engineering and MLOps.
Work
Projects
PatagonIA
Wildfire risk prediction in Argentine Patagonia using ML: web scraping → feature engineering → classification models. End-to-end pipeline in Python.
- Python
- scikit-learn
- Web Scraping
- ML
Breast Cancer Detection
Binary tumor classification model comparing Logistic Regression, Random Forest, and Decision Tree with GridSearchCV. Best result: Logistic Regression with 99.7% ROC AUC, 95.6% accuracy, and 93.8% F1-score.
- Python
- scikit-learn
- GridSearchCV
- Classification
GlorIA
In progressFootball predictor optimized for Expected Value (EV): Monte Carlo simulations + ensemble stacking (XGBoost + neural network). Works for any tournament or league.
- Python
- XGBoost
- PyTorch
- Monte Carlo
FundaData
Automated metrics platform for elderly care, childcare, and day centers. Consolidates institutional metrics through automated workflows.
- TypeScript
- Supabase
- PL/pgSQL
- n8n
Trackealo
Full-stack personal finance app: custom categories, dashboards, and real-time data.
- React
- Vite
- JavaScript
- Supabase
- PostgreSQL
Climate Change ARG
Climate change report for Argentina (2000–2023): web scraping pipeline + data cleaning + interactive Power BI dashboard.
- Power BI
- Python
- Web Scraping
F1 Telemetry
Formula 1 race simulation and telemetry analysis in R, from EDA to final visualizations.
- R
- ggplot2
- EDA
Letterbox Project
Letterboxd-inspired web platform built in Java: MVC architecture, Maven build, and JS/HTML frontend.
- Java
- Maven
- JavaScript
- HTML
Live
GitHub Activity
@aformen9 — updates automatically every 10 minutes.
Contributions
Recent activity
Let's Talk
Got a project or an idea? Reach out.
Always up for projects like yours. Reach out, I reply fast.
info@agustinformenti.dev