Efficient Information Retrieval – Part I

Do you spend too much time searching for information in PDFs, folder structures or internal documentation? Traditional AI chatbots quickly reach their limits here – either because they do not have access to your data or because sensitive content would have to be passed on to external providers.

In part I of this four-part workshop series, you will learn how to set up a local, data-secure system on your laptop. This system enables you to retrieve specific information from your own documents and generate context-related, comprehensible answers from it – without sharing your data with third-party providers.

Typical applications include internal knowledge databases, guidelines, and project documentation.

A retrieval-augmented generation system (RAG system) can generate answers based on your data. In this workshop you will learn the fundamental building blocks: local large language models (LLM), embeddings, and vector databases. By the end, you will understand how RAG works, its advantages over pure LLM queries, and how to evaluate whether RAG is suitable for your own data and use cases.

In the practical part of this workshop, you will install a complete minimal version of a RAG system:

· Local LLM (CPU- or GPU-accelerated), hosted with vLLM

· Simple vector database based on Qdrant

· Example documents indexed using different embedding models

· Retrieval using various semantic search techniques

· Generating answers based on relevant information and your own prompts

You will learn the basics of semantic vectors (e.g., distance metrics such as cosine similarity) and explore various open-source models. Advanced concepts such as Graph RAG or HNSW are introduced conceptually but not yet covered in depth.

👉 This workshop lays the foundation for all other parts of the series.
After the workshop, you will have a working local RAG system and know how to transfer it to your own data.

Requirements

Experience with Python, GitHub, and Docker.
or completion of our Getting Started Guide
or participation in our basic workshop ‘First steps towards your own AI application’.

Registration & participation

Registration is binding. If you are unable to attend, please cancel your ticket in good time to allow other participants to take your place.

Important Information

Photographs, audio recordings and films will be made during the event. By participating in the event, you agree that photos, audio and video recordings in which you are recognizable may be published as part of the public relations work of the AI Service Center Berlin-Brandenburg and the HPI.

The is a project of the Federal Ministry of Research, Technology and Space. Its aim is to lower the barriers for using AI in business and society.

Are you interested in our free workshops and other events that we offer? Then sign up for our newsletter.

Know what’s Happening Next — before everyone else does.

Get App

Host or Publisher KI-Servicezentrum Berlin-Brandenburg @ HPI