Vector database with RAG
Teach LLM to read your documents
- Date
- Time
- 18:00 - 20:00
- Venue
- PEF CZU DII
- Lecturer:
- Sergiu Nica, Štěpán Pešout
How it went
Attendees gained an overview of the relationship between vector databases, RAG and large language models and saw a basic implementation of document querying.
The vector database technical workshop with RAG introduced Retrieval-Augmented Generation as one of the most practical ways to extend language models to work on your own data. Participants went through the entire basic flow: from converting a document into a vector representation, through searching for relevant passages, to generating an answer based on the found context. The quality of the retrieval phase, chunking, relevance and the impact of the architecture on the resulting behavior of the system were discussed. The workshop provided a solid technical foundation for the further construction of knowledge-based AI applications.