Hi, I'm Naia 👋

Data Science & AI · Responsible ML · MLOps

Final-year BSc student at Deusto University (GPA 8.13/10), currently finishing my degree project: a natural-language interface for real-world traffic data using RAG + LLMs. I did an exchange semester at Sungkyunkwan University in Seoul (GPA 4.42/4.5) and spent ~700 hours as an AI Data Annotation Specialist at DataAnnotation, evaluating model safety and bias in English and Spanish.

My work sits at the intersection of machine learning engineering and responsible AI — I care about building models that work reliably in production, not just in notebooks.

---

🔧 What I've built

| Project | What it does | Stack |
|---|---|---|
| StreamGuard | Production ML monitoring — drift detection, SHAP explainability, performance tracking, REST API | FastAPI · SciPy · SHAP · Streamlit · Docker |
| FairLens | AI fairness auditing toolkit — 5 bias metrics, intersectional bias, HTML compliance reports (EU AI Act / EEOC) | Python · scikit-learn · Streamlit |
| [data-analyst](https://github.com/NaiaLorente/ai-data-analyst) | Natural-language data analysis agent | LangChain · LLMs · DuckDB |
| anomaly-detection | Anomaly detection pipeline | Python · scikit-learn |
| Recommendation System | Neural collaborative filtering + hybrid content model on MovieLens | PyTorch · TF-IDF |
| Distributed Big-Data Pipeline | Hadoop/Spark cluster on AWS EC2, Airflow DAGs, 5k+ shooting records | PySpark · Hadoop · Airflow · AWS |
| CNN & Adversarial Robustness | CIFAR-10 classifier with FGSM adversarial attacks and adversarial training | PyTorch · Torchvision |
| Time-Series Forecasting | Bike-share demand with ARIMA, SARIMAX, and LSTM benchmarks | statsmodels · pmdarima · PyTorch |

---

🛠 Skills

ML & AI — supervised/unsupervised learning, deep learning, time-series, NLP, model evaluation, AI ethics & bias detection
MLOps — Docker, FastAPI, GitHub Actions, Streamlit, Grafana
Big Data — PySpark, Hadoop, Airflow, AWS EC2
Languages — Python · SQL · R · Java
Tools — PyTorch · scikit-learn · LangChain · DuckDB · ChromaDB · MongoDB · Neo4j

---

🎓 Education & Certs

• BSc Data Science & AI — Deusto University, Bilbao (2022–2026) · GPA 8.13/10
• Exchange — Computer Science — Sungkyunkwan University, Seoul (2025) · GPA 4.42/4.5
• WorldQuant University — Applied Data Science Lab (2026)
• Google — AI & Productivity (2025)
• Cambridge C1 Advanced (2024)

🌍 Languages

Spanish (native) · English (C1) · French (B2) · Basque (B1)

---

📫 [email protected] · LinkedIn · Bilbao, Spain

Collapse