I'm a Data Scientist and AI practitioner with a background in Computer Science and more than ten years of experience in software development and digital solutions. I build end-to-end data and AI projects spanning machine learning, predictive analytics, natural language processing, computer vision, and generative AI applications.
My portfolio includes customer segmentation, time series forecasting, NLP classification, toxic communication pattern detection, recommendation systems, and computer vision projects, with a strong focus on feature engineering, model evaluation, explainability, and reproducible workflows. I enjoy transforming complex data into actionable insights and designing AI-powered solutions that combine technical rigor with real-world impact.
I am particularly interested in Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, Explainable AI, Large Language Models, Retrieval-Augmented Generation (RAG), Fine-tuning, and the development of responsible AI systems that address meaningful business and social challenges.
- Machine Learning & Artificial Intelligence
- Computer Vision & Deep Learning
- Natural Language Processing & Text Classification
- Large Language Models & Generative AI
- Retrieval-Augmented Generation (RAG)
- Fine-tuning & Transfer Learning
- Responsible AI & Ethical AI Design
- Data Analytics & Visualization
- Predictive Modeling & Time Series
- Unsupervised Learning & Clustering
- Explainable AI & Model Evaluation
- Reproducible Data Science Workflows
Full-stack GenAI application with multi-model AI pipeline, RAG, contextual risk assessment, and Responsible AI design. Built with FastAPI, PostgreSQL + pgvector, fine-tuned DistilBERT, OpenAI GPT-4o-mini, and Streamlit. 🔗 Repo: https://github.com/MapiAI/SafeGuard-AI
Unsupervised learning project using segmentation and clustering to identify customer personas and behavioral patterns.
🔗 Repo: https://github.com/MapiAI/TravelTide-Customer-Segmentation
End‑to‑end forecasting project with MLflow tracking, Streamlit apps, and model evaluation.
🔗 Repo: https://github.com/MapiAI/corporacion_favorita
Image classification on the CIFAR-10 dataset using a custom CNN and a transfer learning approach based on a pretrained ResNet50 model.
🔗 Repo: https://github.com/MapiAI/cifar10-image-classification
Supervised machine learning project that classifies vehicle types using tabular features extracted from vehicle silhouettes.
🔗 Repo: https://github.com/MapiAI/Vehicle-Silhouette-Classification-Project
Exploratory data analysis and visualization on car datasets.
🔗 Repo: https://github.com/MapiAI/Car-Data-Analysis
- FIFA23 Dashboard → Interactive player performance dashboard
- Unicorn Companies Dashboard → Business intelligence visualization of startup growth
- Generative AI application development (RAG, LLMs, fine-tuning)
- Deep Learning & Computer Vision
- Model Deployment and MLOps
- Reproducible AI and Machine Learning Workflows
I’m open to collaborating on data science projects that combine technical depth with clear communication and visualization.
- GitHub: https://github.com/MapiAI
- LinkedIn: https://www.linkedin.com/in/maria-petralia
- Tableau Public: https://public.tableau.com/app/profile/maria.petralia
I find beauty in both nature and numbers. When I’m not analyzing data or creating vector images, you’ll find me designing cakes 🎂, tending to my garden 🌱, or practicing yoga & tai chi 🧘♀️. Creativity and logic are the two forces that shape my world.