Training · Live online · Taught in Italian
Four courses, each in three levels, to understand how AI models work, use them well every day, build them and defend them against attacks. Live classes, exercises on real tools and materials to keep.
Catalog
Each course has three 4-hour levels: beginner, intermediate and advanced. Start from the one that fits you. Dates and fees are being finalised, and everyone who pre-registers gets the calendar first, with no commitment.
What generative AI and language models really are, without the maths.
For anyone who works with AI or has to decide whether to adopt it.
What generative AI is, how an LLM works, what it can do and where it fails
Pre-registerRAG, agents and business use cases
Pre-registerChoosing and assessing an AI project: costs, risks and the EU AI Act
Pre-registerWriting instructions that get good work out of language models, with any assistant.
For professionals, managers and teams who use AI every day.
Anatomy of a prompt: role, context, examples and format
Pre-registerWorking with documents and data, step-by-step reasoning, reusable prompts
Pre-registerChecking and evaluating answers, multi-step workflows and agents
Pre-registerNeural networks from the basics to transformers, with hands-on Python.
For developers, analysts and students with basic Python.
Neural networks, gradient descent and a first model in PyTorch
Pre-registerConvolutional networks and sequence models
Pre-registerAttention and transformers, fine-tuning and deployment
Pre-registerHow LLMs, RAG systems and agents are attacked, and how to defend them.
For technical, security, data science and IT teams.
Threats to AI systems: OWASP Top 10 for LLMs, MITRE ATLAS and the EU AI Act
Pre-registerAdversarial ML, deepfakes, prompt injection and jailbreaking
Pre-registerDefending RAG and agents: guardrails, red teaming and monitoring
Pre-registerAt the intermediate and advanced levels you work on Pandora Lab, a test RAG assistant online: you try the attacks from class on it, then we discuss the defences together.
For companies and public bodies
Every course can run privately for a group of colleagues, online or in person in Rome. We adapt examples and exercises to your tools, your data and your starting level.
Instructor
PhD in Operations Research from Sapienza University of Rome. He has spent almost ten years working on artificial intelligence, language models and agents for companies and public administration. He teaches AI security in master's programmes and runs courses on generative AI and prompt engineering for managers and executives.
He founded DOMIKA Labs, where he builds research tools on open data, and writes the weekly AI Security Research Brief.
Pre-registration
Pre-registration is free and doesn't commit you to anything. We'll write to you when dates and fees for your course are ready.
