Enrico Zanetti

Hi, I'm

AI Engineer based in Trento, Italy

Building AI solutions

View My Projects Get in Touch :)

What I Work With

AI / ML

PyTorch PyTorch TensorFlow TensorFlow Hugging Face Hugging Face LangGraph LangGraph OpenAI OpenAI Anthropic Anthropic MCP MCP

Backend

Python Python Django Django FastAPI FastAPI Celery Celery RabbitMQ RabbitMQ PostgreSQL PostgreSQL Pydantic Pydantic uv uv Ruff Ruff pytest pytest

Frontend

TypeScript TypeScript React React Astro Astro Tailwind CSS Tailwind CSS Vite Vite

DevOps

Docker Docker Linux Linux Git Git GitHub Actions GitHub Actions

Experience

My journey so far.

Terranova Software
work

Jun 2024 - Present

AI Engineer

Terranova Software

Terranova Software · Trento, Italy

  • Develop and maintain WISE, an enterprise AI chatbot powered by LLMs and RAG: full-stack work spanning API design, Celery + RabbitMQ orchestration, PostgreSQL, and React/Vite frontend.
  • Shipped key features end-to-end: SSO authentication, a tool-using AI agent, and parallel queue processing for improved throughput.
  • Research and benchmark LLMs for production use; represent the company at industry events including a talk on AI agents for enterprise automation.
certification

Jan 2026 - Mar 2026

Responsive Web Design

freeCodeCamp freeCodeCamp

freeCodeCamp · Online

  • As part of my path toward becoming a full-stack AI Engineer, I completed this hands-on course to build accessible, responsive web pages with HTML and CSS: the foundation behind this very website.
View Certificate
certification

Mar 2025 - May 2025

AI Agents Course

Hugging Face

Hugging Face · Online

  • Driven by a growing interest in AI agents, I took this course to explore their foundations, from tool use and reasoning to multi-agent orchestration, and built a custom agent as course project.
View Certificate
Stanford University
certification

Sept 2024 - Dec 2024

Natural Language Processing with Deep Learning: XCS224N

Stanford University

Stanford University · Online

  • After graduating, I wanted to deepen my understanding of neural NLP, from word representations and dependency parsing to pretraining transformers, bridging theory and hands-on PyTorch implementation.
Stanford Online Certificate
University of Trento
education

Sept 2022 - Sept 2024

Master's Degree in Data Science

University of Trento

University of Trento · Trento, Italy

  • Driven by a strong passion for machine learning, deep learning, and statistical methods, my studies culminated in a thesis on RAG-based personalization and LLM evaluation for AI chatbots. I also really enjoyed a course on the Internet of Things, where I got to explore the application of AI in robotics.
My Thesis
Utrecht University
education

Aug 2023 - Feb 2024

Erasmus Exchange Program

Utrecht University

Utrecht University · Utrecht, Netherlands

  • During my Master's in Data Science, I had the opportunity to spend a semester in Utrecht. This experience broadened my perspective on data science and AI through advanced courses in machine learning and reinforcement learning within an international academic environment. It remains one of the most rewarding experiences of my life, both academically and personally.
University of Trento
education

Sept 2017 - Feb 2021

Bachelor's Degree in International Studies

University of Trento

University of Trento · Trento, Italy

  • After high school, I followed my interest in geopolitics, history, and writing with a Bachelor's degree in International Studies. Along the way, I discovered an even stronger passion: coding and technology. Still, I consider this experience invaluable: it gave me a multidisciplinary foundation spanning European law, international relations, and quantitative methods that continues to shape how I think about systems, policy, and data.

Projects

From ideas to working software.

Cortex

Cortex is a knowledge base that lets users upload and search internal company documents, then chat with an AI assistant that answers questions in natural language with citations grounded in those documents. Under the hood, documents are parsed, chunked, and embedded for retrieval, and the AI assistant talks to a dedicated MCP (Model Context Protocol) server over HTTP rather than calling the backend directly, ensuring the same tools and interface work identically for the built-in chat and any external AI agent.

Released August 2026 · v0.1.0

MoodStream

Real-time facial emotion recognition module for human-AI interaction. A TFLite CNN classifies expressions from your webcam and streams the emotional state through an IoT pipeline visualized live in Grafana. Designed to run on resource-constrained hardware.

Released May 2026 · v3.0.2

Blog

Contact

Let's build something together :)

Email

ezan@enricozanetti.dev

GitHub

@EnricoZanetti

LinkedIn

Connect with me

Instagram

@ezanico

X

@paranoizan