About me
I'm a Deep Learning Engineer with a strong background in coding and artificial intelligence. My focus is vision-language models and GenAI inference, and I'm passionate about taking that research somewhere it actually gets used.
I studied Computer Science at the University of Modena and Reggio Emilia, with a master's thesis on fine-tuning vision-language models — CLIP and SigLIP — for domain-specific image retrieval. That thesis grew out of an internship at Tetra Pak, and the work didn't stop at the thesis: it's the same family of problems I now work on full time, making those models useful on real industrial data.
Along the way I've led the driverless division of a Formula Student team and published a paper on microservice placement in fog computing. More recently I've been living in Kubernetes and MLOps, which turned out to be the useful half of the job nobody teaches you: if I build a model, I want to own the pipeline that ships it.
skills
- ML / Deep Learning
- PyTorch · Hugging Face(accelerate-transformers-peft) · vLLM · Pandas · NumPy · Scikit-Learn · Plotly · Streamlit
- AI
- VLMs · LLMs · RAG · MoE · Fine-tuning · Agents
- Coding
- Python · C++ · C · Java
- MLOps / Infra
- Docker · Kubernetes · FluxCD · CI/CD pipelines · Azure · AWS
- Databases
- PostgreSQL · MySQL · SQLite · Weaviate · PGVector
- Languages
- Italian (native) · English