// Index Definition: Create an auto-embedding index
await collection.createSearchIndex({
  name: 'product_auto_embed_idx',
  type: 'vectorSearch',
  definition: {
    fields: [
      {
        type: 'autoEmbed',  // New index type!
        modality: 'text',
        path: 'description',
        model: 'voyage-4'
      },
      {
        type: 'filter',
        path: 'author'
      }
    ]
  }
});
// Insert documents (no manual embeddings needed)
await collection.insertOne({
  description: 'Wireless headphones with noise canceling',
  author: 'TechCorp'
});

// ============================================
// Manual embedding generation (No longer needed!!)
// ============================================
// Previously, users had to:
//
// 1. Set up external embedding service
// const openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY });
//
// 2. Generate embeddings for documents
// const docEmbedding = await openai.embeddings.create({
//   model: 'text-embedding-ada-002',
//   input: 'Wireless headphones with noise canceling'
// });
//
// 3. Store embeddings in documents
// await collection.updateOne(
//   { description: 'Wireless headphones with noise canceling' },
//   { $set: { embedding: docEmbedding.data[0].embedding } }
// );
//
// 4. Generate embeddings for queries
// const queryEmbedding = await openai.embeddings.create({
//   model: 'text-embedding-ada-002',
//   input: 'audio equipment'
// });
//
// 5. Use vector in query
// queryVector: queryEmbedding.data[0].embedding
// ============================================

// Query: Use text query instead of vector
const results = await collection.aggregate([
  {
    $vectorSearch: {
      index: 'product_auto_embed_idx',
      path: 'description',
      query: { text: 'audio equipment' },
      numCandidates: 100,
      limit: 5,
    }
  },
  {
    $project: {
      description: 1,
    }
  }
]).toArray();
// Results: Retrieved documents with similarity scores
// [
//   {
//     description: 'Wireless headphones with noise canceling',
//   }
// ]