Skip to content
Nappai Documentation

Supabase Vector Search

This component lets you search through your documents stored in Supabase, a database service. Instead of just looking for exact matches, it understands the meaning of your search terms and finds the closest matches. This is useful for finding similar documents even if they don’t use the exact same words.

Relationship with Supabase

This component directly interacts with the Supabase database to store and search your documents. It uses Supabase’s vector search capabilities, which allow for semantic searches based on the meaning of your data. You’ll need a Supabase account and to have set up a table to store your data.

Inputs

  • Table Name: The name of the Supabase table where your documents are stored. (Advanced setting)
  • Query Name: A name you give to this specific search query. This helps you organize your searches.
  • Search Query: Type in the words or phrases you want to search for. This is where you describe what you’re looking for.
  • Ingest Data: The data you want to add to your Supabase database for searching. (This is usually handled automatically by other components in Nappai).
  • Embedding: This setting defines how your documents are represented for searching. (This is usually handled automatically by other components in Nappai and should be of type “Embeddings”).
  • Number of Results: How many of the most similar documents you want the search to return. The default is 4. (Advanced setting)
  • Credential: Your Supabase API credentials, allowing Nappai to access your Supabase account.

Outputs

The component returns a list of documents that best match your search query. Each document will be presented in a way that’s easy to understand within the Nappai dashboard. You can then use this list of documents in other parts of your workflow.

Usage Example

Let’s say you have a collection of customer support tickets stored in Supabase. You want to find tickets similar to a new ticket you just received. You would:

  1. Enter the name of your Supabase table containing the tickets in the “Table Name” field.
  2. Enter a descriptive name for this search (e.g., “Customer Ticket Search”) in the “Query Name” field.
  3. Type a summary of the new ticket in the “Search Query” field.
  4. Nappai will then return a list of similar tickets from your Supabase database, helping you quickly find solutions or identify patterns.

Templates

[List of templates where the component is used will be populated here. This section will be updated as templates are created.]

  • VectorStoreInfo: Provides information about the vector store being used.
  • Self Query Retriever: Helps generate better search queries automatically.
  • [Other Components]: Many other components in Nappai can be used to prepare data for input into the Supabase Vector Search or to process the output. See the full list of components in the Nappai documentation.

Tips and Best Practices

  • Use clear and concise search queries for better results.
  • Experiment with different search terms to refine your results.
  • Ensure your Supabase table is properly configured for vector search.

Security Considerations

  • Protect your Supabase API credentials. Do not share them publicly.
  • Regularly review the permissions associated with your Supabase account to ensure only authorized users have access.