VoltAgent Qdrant Example

This example demonstrates how to use VoltAgent with Qdrant as the vector database for retrieval-augmented generation (RAG).

Features

• Two agents:
  • Assistant with Retriever: Uses semantic search on every interaction.
  • Assistant with Tools: LLM decides when to search autonomously.
• Tracks sources and references for each answer.

Usage

1. Install dependencies:

   $ docker run -p 6333:6333 qdrant/qdrant
   $ pnpm install

2. Set up environment variables in .env (see .env.example for reference):

   • QDRANT_URL
   • OPENAI_API_KEY (required for embeddings)

3. Start the example:

   pnpm run dev

You can manage your vectors from the Qdrant dashboard.

Example Questions

Try asking questions like:

• What is VoltAgent?
• Tell me about vector databases
• How does Qdrant work?
• What is RAG?
• What is TypeScript?

Notes

• The retriever will automatically populate the Qdrant collection with sample documents if empty.
• Both agents track which documents were used for answers. Check context.get('references') for source IDs and scores.