readyset VS MeiliSearch

- Some platforms like Supabase Realtime [3] and Firebase offer subscription models to database changes, but these solutions fall short when dealing with complex queries involving joins or group-bys.

My vision is that the modern frontend to behave like a series of materialized views that dynamically update as the underlying data changes. Current state management libraries handle state trees well but don't seamlessly integrate with relational or graph-like database structures.

The only thing I can think of is to implement it by myself, which sounds like a big PITA.

Anything goes, Brainstorm with me. Is it causing you headaches as well? Are you familiar with an efficient solution? how are you all tackling it?

[1] https://readyset.io/

Postgresql + MySQL Cache https://github.com/readysettech/readyset

I just wanted to give a high five for having Jepsen tests for this: https://github.com/readysettech/readyset/tree/stable-240117/...

This example is a great use case for partial incremental view maintenance systems like ReadySet: you automatically get something like the “prepopulating the cache” section (toward the end of the blog) while only caching the data the application is using, and avoiding the need to manually implement any sort of invalidation logic.

(Disclaimer: I used to work for them, but don’t anymore. It’s all available for free on GitHub though for anyone interested: https://github.com/readysettech/readyset)

There are systems that will do that for you like https://readyset.io/.

ReadySet

readyset.io is the company that jonhoo was associated with for work on noria

Might be interesting to try out https://readyset.io for this use case.

On the caching topic, I found the ReadySet(né Noria) approach to be extremely interesting.

We needed to introduce a new service for search. As we settled on using meilisearch, we needed a way to sync updates on our models with the records in meilisearch. We could've continued to use callbacks but we needed something better.

Meilisearch

In this case, a good strategy is to use vector search only when the keyword/prefix search returns none or just a small number of results. A good candidate for this is MeiliSearch. It uses custom ranking rules to provide results as fast as the user can type.

With the VertexAI endpoint set up and tested, our next step is to work with LibreChat. LibreChat is an open-source ChatGPT clone that can integrate with various AI models, including the PaLM 2 models via the VertexAI API. It's built using React, MongoDB, and Meilisearch technologies.

Meilisearch seems like it is the best open source option.

https://www.meilisearch.com/

Starting with v1.3, you can use Meilisearch as a vector store. Meilisearch allows you to store vector embeddings alongside your documents conveniently. You will need to create the vector embeddings using your third-party tool of choice (Hugging Face, OpenAI). As we published the first v1.3 release candidate, you can try out vector search today.

I work at Meilisearch, an open-source search engine built in Rust. 🦀

https://www.meilisearch.com/ https://github.com/meilisearch