MongoDB: Vector Search, Displacing Postgres & DynamoDB

This is a snippet of the transcript, sign up to read more.

Moving on to a few questions about MongoDB specifically, coming in from the AI side. The first one is, does it make sense? What's your view of MongoDB's vector search capabilities and packaging that in? Maybe this is probably too recent, or definitely too recent for you to have encountered while you were at MongoDB, but do you have a view on those new products now?

If I wanted to use the best vector data store, I might have to use Pinecone or Weaviate, and then keep it in sync with the metadata saved in a different database. There's always data movement and data plumbing when using various point solutions. But MongoDB doing it all under one umbrella means I can move faster. It may not be the best product for a vector database, but the integrations and the fact that it's all connected well means I have significantly less engineering work to do. As a startup, for any company dealing with this for the first time, I think that's a better way to iterate and try things much faster. Until one day, maybe I figure out I really need a very powerful vector store, and at that point, I would move away from MongoDB.

This is a snippet of the transcript, sign up to read more.

In terms of the internal MongoDB narrative of going after the relational world, did you find that MongoDB was successful in displacing relational use cases? Or was MongoDB just benefiting mostly from the rise of new applications that found it easier to start on NoSQL? Was MongoDB actually successful at taking out Postgres and other relational databases?

When it came to Postgres, I don't think we were replacing it as fast as we were replacing Oracle and mainframes. So there was a lot of growth there. And then the other growth was with high-tech companies like Box, Unity, PayPal, Rivian, Tesla. All these companies are building everything on MongoDB.

This is a snippet of the transcript, sign up to read more.