Hello everyone,
im having alot of troubles to filter the data that retrives from every table in my supabase..
all the data that i need to retrive from supabase is in tables and not in files so when i tried RAG supabase from youtube tutorials i didnt succeed…
what am i missing is someone can please help me with that problem?
Hello davidlevi13, i’ll try to help yaa.
If your data is already in Supabase tables, you dont need RAG just to filter. RAG is only useful when you want semantic search (like “find texts that talk about X, even if the keyword is different“).
Maybe there are several suggestion that i can share:
- If you just want exact filters (by id, date, user_id, etc.), use the Supabase node in n8n:
- Operation: Get All
- Add your filters, for the example: user_id eq {{$json.userId}}.
- Select only the columns you need, dont select all.
- Add limit and order if you want.
- If you really need semantic search (similar meaning, not exact match), then you’d set up pgvector in Supabase. That means you need to create an embedding column in your table, store OpenAI embeddings for each row, and search using cosine similarity.
So.. if all you want is filtering rows, just skip RAG. Just query the table directly with filters. Use RAG only if you need semantic matching.
I hope this will solve your problem. 
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.