Integrations

Weaviate (vector search)

Use Weaviate as a vector provider for the Semantic Search built-in tool. Add Semantic Search to an Agent or Flow to retrieve content from a Weaviate collection. Weaviate is a provider option, not a separate search tool. The tool also supports the built-in pgvector store and Cloudflare Vectorize.

What is Weaviate?

Weaviate is an external vector database that supports semantic search. Use a Weaviate collection to retrieve content for retrieval-augmented generation (RAG).

Choose a vector provider

Choose a provider based on where you store the vectors:

  • Choose pgvector when you store searchable data in Runtype Records.
  • Choose Weaviate when you store searchable data in a Weaviate collection.
  • Choose Cloudflare Vectorize when you store searchable data in a Vectorize index.

Before you begin

Create a Weaviate cluster and collection, and collect the cluster URL and API key. Configure an OpenAI API key in Settings > Models. Semantic Search uses OpenAI to generate query embeddings.

Set up Weaviate

To connect a Weaviate instance to Runtype, follow these steps:

  1. Create a cluster in Weaviate Cloud Console.
  2. Copy the cluster URL and API key from Weaviate.
  3. In Runtype, open Settings > Integrations.
  4. In Vector Databases, click Add Configuration for Weaviate.
  5. Enter a configuration name, a Weaviate Instance URL, and an API Key.
  6. If you have more than one Weaviate configuration, select Set as default Weaviate configuration. The Semantic Search tool uses the default configuration.
  7. On the Test Connection tab, click Test Connection and confirm that the connection succeeds.
  8. Click Save Configuration.

After you connect Weaviate, add Semantic Search to an Agent or Flow. Configure the Weaviate-backed search instance with these settings:

  1. Add the Semantic Search tool to an Agent or Flow.
  2. Select Weaviate as the vector provider.
  3. Enter the name of the Weaviate collection that contains the documents that you want to search.
  4. Enter a Display name for the search instance.
  5. Set Similarity threshold between 0 and 1. Lower values return more results. Higher values return only closer matches.

You can add multiple Semantic Search instances to the same Agent or Flow. Point each instance to a different Weaviate collection.

RAG with Weaviate

Build retrieval-augmented generation (RAG) by adding Semantic Search with Weaviate as the provider to an Agent. In a Flow, add a Vector Search step and select Weaviate as the provider. Pass the search results to the step that generates the answer.

Next steps

Use these pages to continue: