Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Postgres vs VectorDB in 2 minutes.

19,612 görüntüleme • 1 ay önce •via X (Twitter)

9 Yorum

RK profil fotoğrafı
RK1 ay önce

For most enterprise knowledge bases today, Postgres with pgvector is enough. Don't spend more.. Cost rule: pgvector <$200/mo up to 50M. Pinecone/Weaviate/Qdrant = $500–$2000+ at 100M+. Scale decides, not hype.”

Priyanka Vergadia profil fotoğrafı
Priyanka Vergadia1 ay önce

Absolutely!! But most enterprises run towards what is most talked about and invest in stuff they don't need :)

Paul Alvarado profil fotoğrafı
Paul Alvarado1 ay önce

Thanks for the great explanation. It help me to understand embeddings and vector DBs in general.

Rakesh profil fotoğrafı
Rakesh1 ay önce

Thank you, pgvector is useful for Millions of records, easy to setup & use.

Raj C profil fotoğrafı
Raj C1 ay önce

For most enterprises Postgres is good enough when it comes to handling real time transactions and it also supports the vector data; hence you have the expand ability to allow other functions including applying LLM. Great details @pvergadia

Priyanka Vergadia profil fotoğrafı
Priyanka Vergadia1 ay önce

Agree! 👍

Shivaraj profil fotoğrafı
Shivaraj1 ay önce

Is there something called hair db.

zen_winds profil fotoğrafı
zen_winds1 ay önce

Maybe you should try Google AlloyDB it’s worlds best Postgres and also a fastest Vector Database. It even supports 10B+ Vectors scale #AlloyDB

Priyanka Vergadia profil fotoğrafı
Priyanka Vergadia1 ay önce

lol I have more than tried AlloyDB and @andigutmans and team have done an amazing job!

Benzer Videolar