I ship the whole data pipeline. Six projects below across vision-language fine-tuning, streaming engineering, time-series forecasting, and geospatial BI.
About
Most data work splits people into camps: the analyst writes the SQL, the scientist trains the model, the engineer ships the pipeline, the BI lead stands in front of leadership. I do all four.
Florence-2 fine-tuned end to end on a single Colab GPU and shipped as a public web demo. A Kafka and PySpark streaming pipeline carrying daily OHLCV data across 55 tickers. An eleven-model ML tournament against a 10.67:1 class imbalance where the champion caught every single rare event in the test set. A geospatial analysis that joined Fater S.p.A.'s proprietary sales with ISTAT census data, then went up on a screen in front of company leadership. The jury picked the work for individual recognition.
What I'm after: a team that values clarity over cleverness and ships to real users. Strong in Python, advanced SQL (CTEs, window functions, query optimisation), PyTorch, HuggingFace Transformers, Apache Kafka, PySpark, Tableau, and Power BI. My MSc thesis at the University of Naples Federico II focuses on human-robot interaction with reinforcement learning.
Naples-based. Data Analyst, Data Scientist, or ML Engineer roles in Italy and remote across the EU. English C1.
Six projects
Ordered by what most recruiters ask about first. Each card opens a deeper write-up with metrics, code links, and an interactive demo where one exists.
We fine-tuned a 771M-parameter vision-language model on 150K image-question pairs and shipped it live on HuggingFace Spaces.
We built an Apache Kafka + PySpark MLlib pipeline streaming daily OHLCV across 55 per-ticker topics, then clustered with KMeans and PCA.
We built a GitHub GraphQL extractor feeding a Gemini LLM that infers technical skills, project archetypes, and seniority signal for any handle.
We ran an eleven-model tournament across regression and classification; Random Forest reached F1 0.667 and recall 1.00 on a heavily imbalanced minority class.
I trained a custom 1D CNN that edged out SARIMA, ARIMAX, and Prophet on the Open University click-stream forecast (MAE 0.199, MAPE 1.9%).
We joined Fater proprietary sales with ISTAT census in MySQL and ranked districts by per-capita store potential. I presented the work solo to Fater leadership.
Currently building
Active work, mid-stride. Two on reinforcement learning, one on safety-conscious learning in human-robot interaction. Click any card to see the current status and visuals.
A simulated warehouse where six autonomous forklifts learn to pick pallets and avoid each other. Three RL methods benchmarked against three classical path-planning baselines under rigorous statistical analysis.
An RL agent that learns to find objects in natural images by iteratively refining a bounding box through geometric actions, built on frozen CLIP features. A reimplementation-with-modern-components of Caicedo and Lazebnik (ICCV 2015) on Pascal VOC 2007, not a strict reproduction.
MSc thesis. A web-based HRI study comparing four shielding conditions for a Q-learning agent on a 7x7 grid: no shielding, standard preference shielding, Adaptive Shielding (confidence gate), and Hard/Soft per-object Shielding. Participants will watch the agent navigate, express directional preferences, and answer questionnaires.
HypothesisDoes adding a confidence gate (Adaptive Shielding) or a Hard/Soft per-object enforcement split to the existing Preference Shielding mechanism improve how transparent and trustworthy a learning robot looks to a human observer, without slowing down how quickly it learns the task?
Credentials
Federico II in partnership with Nokia, TIM, and PagoPA
Postgraduate programme on 5G and digital transformation. Industry-aligned curriculum delivered jointly with EU telecom and public-tech partners.
VisitUniversità degli Studi di Napoli Federico II, Apple Developer Academy
Signed by Giorgio Ventre, Scientific Director of the Apple Developer Academy at Federico II.
View certificateFater S.p.A. in collaboration with the MSc Data Science programme, Federico II
Attendance signed by Fater's Sales & Digital Business Analyst Manager, Head of Data & Analytics, and Sales & Digital Data Scientist Project Manager.
View certificate