Hi, how can i implement natural language search with typesense in Python? Also, I'd like to use ollama models and openai api for LLMs in the natural language search. Please suggest related solutions @Kishore Nallan

Have you checked the docs on vector search? It covers these details.

I'm referring to such a functionality in python https://typesense.org/docs/guide/natural-language-search.html#use-case

Ah, got it. We don't have a python / openai based workflow for this yet.

Any workaround in typesense so that i could engineer this myself?

The approach is just using an LLM to construct a structured Typesense query from a natural language query, based on some examples in the prompt. You should be able to adapt this approach to OpenAI.

Can you suggest a way to integrate ollama/huggingface models with this approach?

I doubt that the smaller models will be able to do this reliably

I get it i just want to try it, any idea on how to integrate ollama in this one??

Right now, it's experimental. For using ollama you have to find a library that replaces the interactions with Gemini API with ollama API calls.