Dragwyb Docs Dragwyb Docs
Visit Website
Updated on Oct 17, 2026

Configure SQLite Vector Storage

Select Dragwyb Click to Chat’s SQLite (Local) vector-storage option, configure embeddings, index content, and test retrieval.

The SQLite (Local) option is the built-in local vector-storage choice in Dragwyb Click to Chat. It avoids configuring an external vector database service. The current PHP adapter stores embeddings in plugin-managed WordPress database tables, so “local” refers to storage on your WordPress installation rather than a separate SQLite file.

Choose local vector storage

  1. Open Click to Chat → AI Assistant → Knowledge Base → Database & Indexing.
  2. Under Database Storage Engine, select SQLite (Local).
  3. Under Embedding Configuration, select OpenAI (text-embedding-3-small) or Google Gemini (gemini-embedding-001). Configure the corresponding provider API key in the AI Engine & Prompt settings before indexing.
  4. Save the Knowledge Base settings.
  5. Select Index Content Now to process the selected website content and other configured sources.
  6. Wait for any queued indexing and embedding work to finish, then test with a question answered by the indexed content.

When to troubleshoot

  • No relevant answer: Confirm the right post types and sources are selected and that indexing completed.
  • Embedding errors: Check the selected provider and its API key, then review the plugin’s available AI error information.
  • Slow or incomplete results: Test with a smaller, clearly defined question and verify that the answer appears in the source content.

SQLite storage does not remove the need for an embedding provider: the selected provider converts source text into vectors used for similarity search. For a cloud vector database instead, see Connect Pinecone Vector Storage. For the full workflow, see Knowledge Base and Retrieval.