MSc student in Artificial Intelligence Systems @ University of Trento (BSc in Computer Engineering).
I build ML and NLP systems — multi-agent, semantic retrieval, LLM tooling — and I care about making AI more interpretable, sustainable and socially useful.
- 🎓 Studying AI Systems at UniTN (Trento, Italy)
- 🔭 Currently building LLM-assisted tooling for knowledge work (Obsidian + AI pipelines)
- 🌱 Focus: interpretable & trustworthy AI (XAI) · AI for social & environmental impact · knowledge graphs · decentralised systems
- 💬 Open to internships and junior roles in AI/ML and software engineering
- 📫 Reach me: nilo.netti2003@gmail.com
Languages — Python · JavaScript / TypeScript · Java · C / C++ · SQL ML / AI — PyTorch · scikit-learn · embeddings & semantic search (sentence-transformers, FAISS) · Transformers / LLMs · multi-agent systems NLP & Knowledge — NLP · retrieval / RAG · RDF / SPARQL · knowledge graphs (ESCO, Oxigraph) Backend & Tools — FastAPI · Node.js · Git · Linux · Docker
| Project | What it is | Stack |
|---|---|---|
| Autonomous Delivery Agent (ASA) | Hybrid BDI + LLM autonomous multi-agent for the Deliveroo challenge — A* + PDDL planning, LLM strategy, socket coordination. | JavaScript · Node.js |
| Obsidian AI Pipeline | A "second brain" toolkit: semantic (embeddings) note search, an AI-news watcher on a systemd timer, and LLM skills that turn books into notes + Anki flashcards. | Python · LLM |
| Conversational Clustering — BlaBlaClust | Full-stack ML system that clusters conversational text: FastAPI backend, embedding models, UMAP visualisation and evaluation against baselines. (team project) | Python · TypeScript |
| Compositional Image Retrieval (CLIP) | Learned projection head on frozen CLIP for face retrieval with +/− text conditions — Recall@10 from 15.9% to 37.0% on CelebA, with paired-bootstrap evaluation. (team project) | Python · PyTorch |
| Course → Skill Matching (TSM × ESCO) | Maps training courses to the ESCO skills taxonomy with multilingual embeddings, FAISS retrieval and LLM re-ranking. Public AI Challenge with TSM. (team project) | Python · NLP |
Pinned repositories below are kept up to date.
Interested in working together? Let's talk.