Software Engineer graduated from the Universidad de Sevilla, currently looking for my first junior role. I spent this year interning at NTT DATA, building Angular/TypeScript UI on top of Java/Spring Boot services, and finished a Bachelor's Thesis on depression and suicidal-ideation detection (BERT/SBERT) that got a 10/10.
Outside of coursework I taught myself computer vision from zero by building a volleyball tactical-analysis pipeline (YOLOv8 + ByteTrack), and I've got a few other independent projects below — a reinforcement-learning agent, a generative-art tool driven by live weather data, and some smaller web apps.
I like taking things past the notebook stage: training a model is only half the work, the other half is wrapping it in something that actually runs.
Looking for: a junior Software Engineering or ML/NLP role — open to pretty much anything at this point. Full portfolio → norapfr.github.io
Languages
Web & Backend
Data
ML / AI
Tools
Volley Predictor · sports analytics, independent project
ML system predicting international volleyball results: Elo ratings + CatBoost with walk-forward validation, trained on 6,626 historical matches across 158 national teams. Predicts winner, most likely set score, and point margin through a Streamlit dashboard. Winner accuracy: 72.5% (men), 76.0% (women).
Python CatBoost FastAPI Streamlit
SocialMindScan · Bachelor's Thesis, graded 10/10
NLP system for detecting depression and suicidal ideation in text. Compared six model families (SVM, KNN, Random Forest, CNN, BiGRU, BERT/SBERT) across two Reddit-derived datasets, then shipped the best ones behind a Django + React app paired with a support chatbot. Best results: F1 0.893 (CNN + SBERT) for depression, F1 ≈0.909 (SVM + TF-IDF) for suicidal ideation. Companion app: SocialMind_APP.
Python PyTorch BERT/SBERT Django React
VolleyVision AI · computer vision
End-to-end tactical analysis system for volleyball: YOLOv8 for player/ball detection, ByteTrack for multi-object tracking, packaged behind FastAPI + Docker with MLflow for experiment tracking. Models are public on HuggingFace Hub. Built with zero prior CV experience going in.
Python YOLOv8 ByteTrack FastAPI Docker MLflow
LunarEntorno · reinforcement learning
A Deep Q-Network agent (experience replay + target networks) trained on OpenAI Gym's LunarLander. Lands successfully in 85% of evaluation episodes.
Python PyTorch OpenAI Gym
atmospherica · generative art
Turns six real-time atmospheric variables (OpenWeatherMap, ERA5) into generative visuals through a parametric grammar. Fully deterministic — reproducible runs via a seeded XORShift RNG.
Python OpenWeatherMap ERA5
Software Developer Intern — NTT DATA · Jan 2026 – Aug 2026 Built and shipped Angular/TypeScript UI components wired into Java/Spring Boot backend services for public-sector client platforms, and wrote reusable Spring Boot + Hibernate business logic backed by SQL. Also touched some Node.js services. Worked sprint-based with Jira, code quality checked through SonarQube.
B.Sc. Computer Engineering (Software Engineering) — Universidad de Sevilla, 2022–2026 GPA 8.41/10, with distinction in Infinitesimal & Numerical Calculus, Linear & Numerical Algebra, and Computer Networks. Thesis: Artificial Intelligence in Mental Health, graded 10/10.
Also part of the university's Algorithmics Club, competed in Ada Byron Regional 2025.
Languages: Spanish (native) · English (B2, Trinity)
