i’d love to use LLM context filters for cast/tweet search

for example, when I search for a word, I want to provide some context about what I’m looking for, without using those exact words in the actual search

then, an LLM would classify every item found in the search results and show me only the ones that match the intended meaning

Could use RAG for this. Do a comparison of embeddings. 

Would be a fun project to work on - I'd be down to help

mago hacker 🇧🇷🏴‍☠️ building /gnars and /nounspace

Building /nounspace - the digital home for communities | Founding Member /atxdao | Creating community and good vibes wherever I go.

yeeaahh, I know about rag, but I’ll need to rag the next 15 docs, then the tailwind, then the react, and on and on

I don’t like too much generalist architecture, maybe we need more “microservices” with mini models

Could use RAG for this. Do a comparison of embeddings. 

Would be a fun project to work on - I'd be down to help

https://medium.com/@shravankoninti/mastering-rag-a-deep-dive-into-embeddings-b78782aa1259