Description

A two-day hands-on workshop where you build from scratch a system that allows AI to answer questions based on your company documents. RAG is a solution to the AI “hallucination” problem – the model gets only verified facts from your sources, it does not make up information. The first day is about fundamentals: how to process documents, create embeddings and build vector bases. The second day was all about practice: orchestrating the entire pipeline, optimizing the relevance of the results and implementing the finished Q&A application. It’s the link between AI theory and a real-world product.

For whom

Developers, Data Scientists, AI Architects, Data Engineers.

What you gain

You will master the complete process of building a RAG system – from processing PDF and DOCX documents, to creating embeddings, to implementing semantic search. You will learn to select and configure vector databases (ChromaDB, Pinecone, FAISS), optimize chunking and re-ranking strategies for better relevance. You will build a working Q&A application ready to be expanded with your data. You will also receive complete source code and reference architecture for implementation in production.

Price: 32 000 zł net