Hands-On RAG: Build Your Own Document Q&A System

Felix Boelter is an AI Engineer at the Hasso-Plattner-Institut, where he works in the AI Service Center Berlin-Brandenburg. He builds LLM inference infrastructure, develops Retrieval-Augmented Generation (RAG) and agentic AI systems with public-sector and SME partners, and teaches these technologies through hands-on workshops. He is joined by Mario Tormo Romero, an AI Engineer and Senior Data Scientist specialising in NLP, RAG systems, vector databases, and MLOps. Together, they bring extensive experience across the AI and data-science lifecycle.

Retrieval-Augmented Generation enables language models to answer questions using information retrieved from a defined collection of documents. This workshop offers a practical introduction to how a RAG system works—from document ingestion and retrieval to conversational question answering—and how its results can be evaluated. The workshop uses an open-source stack, giving participants a transparent and self-hostable foundation for their own projects.

What participants will do

Participants will work through the main components of a RAG pipeline and build a document question-answering system. They will explore document ingestion with tools such as Docling, connect retrieved content to an LLM-based chatbot interface, and gain a basic understanding of RAG evaluation. The workshop is designed so that participants can understand and continue using the template after the event.

What to expect

Participants can expect a hands-on technical session with a practical outcome: a RAG template that they can take home and build on. No previous experience with RAG is required, but participants should have basic Python knowledge and bring a laptop with Python 3.10 or later and Docker installed. Before attending, they should complete the setup guide: https://github.com/aihpi/workshop-getting-started/

Back to overview