Hi, I’m Lorenzo Sciarretta 👋

I’m a Master’s student in Computer Science at the University of St. Gallen and ETH Zurich.

I have a strong technical foundation with expertise in machine learning, deep learning, and software engineering. Currently, I am conducting research on Weight Space Learning at the AI/ML Lab at HSG with Prof. Damian Borth.

Previously, I worked as a Research Assistant at the ETH AI Center, under the supervision of Prof. Shih-Chii Liu, and as a Quantitative Researcher at the Financial Econometrics Chair (Math & Statistics Department), under the supervision of Prof. M.R. Fengler.

Complementing my technical background, my education at HSG has provided me with a solid understanding of business and strategic management, enabling me to bridge the gap between technical innovation and organizational objectives.

---

Projects

Simple Local RAG

View Project

A very simple Retrieval-Augmented Generation (RAG) system running locally, integrating vector search and prompt chaining for document-based question answering.

LangChain, Python, LLM, tokenizers

SWE Project: Job Broker System

View Project

An auction-based job assignment system using hexagonal architecture. Features job scheduling, worker credit systems, and fault tolerance.

Microservices, distributed systems, Hypermedia ecosystem

CODEX Naturalis App

View Project

Full-stack Java application simulating a board game, covering the full software architecture. Built the network layer (RMI and Socket), designed and implemented game logic, and created TUI and GUI.

Computer Vision: Kidney and Kidney Tumor segmentation with CNN

View Project

Biomedical Computer Vision research project at NECSTLab. I developed a Kidney and Kidney Tumor segmentation software using Deep Learning techniques. The structure of the project is: creation of a custom dataset (from NIFTI images), preprocessing, tuning and training the model

Computer Vision: Unsupervised Deep Learning Registration Framework for Histology Samples with Varied Staining

View Project

Implementation of an Unsupervised Deep Learning Registration Framework for Histology Samples with Varied Staining

Paper: Read Paper

---

Skills

• LLMs: prompting, fine-tuning, security & reliability, architecture optimization, efficient attention mechanisms
• Machine Learning: DFM, SVM, ... (Python, R) Scikit-learn, pandas, Numpy)
• Time-Series Forecasting (Python, R, Julia)
• Deep Learning: LSTM, RNN, NN, Computer Vision, PCA (TensorFlow, PyTorch, Nibabel)
• Strategic Management and Business Integration
• Software Engineering (Java, C)
• Data Visualization (Matplotlib, Plotly, ggplot2, latex, Tableau)
• Databases (SQL, snowlflake)