Skip to content
Nappai Documentation

Couchbase

This component allows Nappai to search and store information using Couchbase, a powerful database system. It’s used when you need to quickly find similar pieces of information within your data.

Relationship with Couchbase

This component directly interacts with a Couchbase database, specifically using its vector search capabilities. This means it can find information based on how similar it is to what you’re looking for, not just by matching exact keywords.

Inputs

  • Bucket Name: The name of the specific storage area within your Couchbase database. Think of it like a folder in a filing cabinet. Required.
  • Scope Name: A further subdivision within your Couchbase bucket. Like a subfolder within a folder. Required.
  • Collection Name: The specific collection of data within the scope. This is where your actual information is stored. Required.
  • Index Name: The name of the search index used to speed up searches. This is automatically created and managed by the component. Required.
  • Search Query: The information you’re looking for. Nappai will find similar items based on this query. Optional.
  • Ingest Data: The data you want to add to the Couchbase database. This is a list of items you want to store. Optional.
  • Embedding: A representation of your data in a format that Couchbase understands for efficient searching. This is handled automatically by Nappai. Required.
  • Number of Results: How many similar items you want Nappai to return in a search. Defaults to 4. Advanced Option.
  • Credential: Your login information for accessing your Couchbase database. Required.

Outputs

The component doesn’t have direct outputs in the traditional sense. Instead, after a search, the results (similar items found) are used directly within Nappai’s workflow. These results are presented to you in a user-friendly format within the dashboard.

Usage Example

Imagine you have customer reviews stored in Couchbase. You want to find reviews similar to a new review you just received. You would:

  1. Provide the correct Bucket Name, Scope Name, Collection Name, and Index Name for your review data.
  2. Enter the new review text in the “Search Query” input.
  3. Nappai will search Couchbase and display similar reviews in the dashboard.

Templates

[List of templates where the component is used – This section needs to be populated with the actual template names from the provided list. For example:]

  • Template A: Uses Couchbase to search customer feedback.
  • Template B: Uses Couchbase to find similar product descriptions.
  • VectorStoreInfo: Provides information about the vector store used by this component.
  • Self Query Retriever: Helps generate search queries automatically.
  • Other Vector Stores (e.g., PGVector, Upstash, Chroma DB): Alternative components that also perform vector searches, but use different database systems.

Tips and Best Practices

  • Make sure your Couchbase database is properly configured before using this component.
  • Use descriptive names for your buckets, scopes, and collections to keep your data organized.
  • Experiment with different search queries to refine your results.

Security Considerations

  • Protect your Couchbase credentials carefully. Do not share them with unauthorized individuals.
  • Regularly review and update your Couchbase security settings.