Preview

Preview Feature — This feature is currently in preview and under active development. APIs and functionality may change. We recommend testing thoroughly before using in production.

Vector Databases

Vector databases store and query high-dimensional vectors for similarity search, enabling powerful AI and machine learning features like semantic search, recommendations, and retrieval-augmented generation (RAG).

When to Use Vector Databases

Use Vector Databases when you need:

• Semantic Search - Find similar content by meaning, not just keywords
• Recommendation Systems - Suggest items based on similarity
• RAG (Retrieval-Augmented Generation) - Give AI agents context from your data
• Duplicate Detection - Find near-duplicate content
• Image/Audio Similarity - Match media by content
• Anomaly Detection - Find outliers in high-dimensional data

---

Supported Providers

Provider	Type	Best For
Pinecone	Managed	Production workloads, scalability
Qdrant	Self-hosted/Cloud	Flexibility, filtering
Weaviate	Self-hosted/Cloud	Hybrid search, GraphQL
ChromaDB	Embedded	Development, prototyping

---

Quick Example

Here's how to set up and use a vector database:

import Ductape, { VectorDBType, DistanceMetric } from '@ductape/sdk';

const ductape = new Ductape({
  user_id: 'your-user-id',
  workspace_id: 'your-workspace-id',
  private_key: 'your-private-key',
});

// Create a vector database configuration
await ductape.vector.create({
  product: 'my-product',
  name: 'Document Embeddings',
  tag: 'doc-embeddings',
  dbType: VectorDBType.PINECONE,
  dimensions: 1536,
  metric: DistanceMetric.COSINE,
  envs: [
    {
      slug: 'dev',
      endpoint: 'https://dev-index.pinecone.io',
      apiKey: 'your-dev-api-key',
      index: 'documents-dev',
      namespace: 'default',
    },
    {
      slug: 'prd',
      endpoint: 'https://prod-index.pinecone.io',
      apiKey: 'your-prod-api-key',
      index: 'documents-prod',
      namespace: 'default',
    },
  ],
});

// Connect to the vector database
await ductape.vector.connect({
  product: 'my-product',
  env: 'dev',
  tag: 'doc-embeddings',
});

// Insert vectors
await ductape.vector.upsert({
  product: 'my-product',
  env: 'dev',
  tag: 'doc-embeddings',
  vectors: [
    {
      id: 'doc-1',
      values: [0.1, 0.2, 0.3, /* ... 1536 dimensions */],
      metadata: { title: 'Introduction to ML', category: 'tutorial' },
    },
    {
      id: 'doc-2',
      values: [0.2, 0.3, 0.4, /* ... 1536 dimensions */],
      metadata: { title: 'Advanced Deep Learning', category: 'advanced' },
    },
  ],
});

// Query for similar vectors
const results = await ductape.vector.query({
  product: 'my-product',
  env: 'dev',
  tag: 'doc-embeddings',
  vector: [0.15, 0.25, 0.35, /* ... query vector */],
  topK: 5,
  filter: { category: 'tutorial' },
  includeMetadata: true,
});

console.log('Similar documents:', results.matches);

---

How It Works

1. Configure - Register vector database with environment-specific credentials using ductape.vector.create()
2. Connect - Connect to the vector database for your environment using ductape.vector.connect()
3. Upsert - Store vectors with IDs and metadata using ductape.vector.upsert()
4. Query - Find similar vectors using cosine, euclidean, or dot product similarity using ductape.vector.query()
5. Filter - Narrow results with metadata filters

Embedding → Vector DB → [Store with metadata]
                     ↓
Query Vector → [Similarity Search] → Top-K Results

---

Distance Metrics

Metric	Description	Use Case
cosine	Angle between vectors (default)	Text embeddings, normalized vectors
euclidean	Straight-line distance	Image features, spatial data
dotproduct	Dot product similarity	When vectors are normalized

---

Next Steps

• Getting Started - Set up your first vector database
• Querying - Advanced query patterns
• Metadata Filtering - Filter results by metadata
• Using with Agents - Connect vectors to AI agents

See Also

• Agents - Build AI agents with vector memory
• Databases - Traditional database operations