Agentic AI / Software Engineer focused on document intelligence, RAG, Python backends, and dependable AI systems.

I build software that turns messy real-world inputs into structured, testable workflows. I completed a Higher Technical Diploma in AI & Data (EQF Level 5) with a final grade of 107/110. My work spans LLM/RAG systems, agentic workflows with LangChain/LangGraph, computer vision and reinforcement learning, plus production-minded software engineering.

• italianRAG — engineering case study
  Evidence-first document intelligence for Italian corporate PDFs, combining typed extraction, Qdrant retrieval, cross-document analysis, and reproducible evaluation. Built during my AI & Data internship at InfoCamere.

• Pefforza 4
  Connect Four AI system combining physical-board computer vision, PPO reinforcement learning, bitboard alpha-beta search, an exact solver, automated evaluation, and CI.

• KeeFetch
  A .NET plugin for KeePass with concurrent favicon retrieval, ranked provider selection, privacy controls, automated tests, and packaged releases.

• ThystTV
  Android/Kotlin media-client work focused on player UX, floating chat, local statistics, release engineering, and upstream maintenance.

• ScrapingStore
  Python data pipeline for scraping, cleaning, storing, and analysing web data.

Python · FastAPI · Pydantic · SQL · Qdrant · RAG · LangChain · LangGraph · LLM/VLM evaluation · pytest · Docker · GitHub Actions · C# / .NET · Kotlin / Android