The magic of vectors with Redis and LLMs
Brian Sam-Bodden and Raphael De Lio from Redis talk with Tjaž Eržen about embeddings, vector databases and using Redis to build LLM and RAG systems that are faster and cheaper than calling a model for every task.
People
Listen
Summary
Tjaž Eržen je ob robu Make IT 2025 gostil Brian Sam-Bodden (Principal Applied AI Engineer) in Raphael De Lio (Developer Advocate) iz Redis.
V pogovoru smo se poglobili v vektorje, vektorske baze, in uporabo slednjih z Redisom pri gradnji sodobnih LLM / RAG - Retrieval-augmented generation sistemov.
Raziskovali smo kako lahko uporaba embedingov in vektorjev pohitri, kot tudi poceni sodobne "AI sisteme".