Quickstart

Go from zero to a fully managed AI-native database in under 5 minutes.

1

Install the SDK

Pulsbase provides native SDKs for TypeScript, Python, Go, and Rust. For this guide, we'll use TypeScript.

Terminal
npm install @pulsbase/sdk
2

Initialize the Client & Provision a Cluster

Grab your API key (`pb_live_...` or `sk_live_...`) from the dashboard and initialize the client. You or your AI Agents (Cursor, Claude, AutoGPT) can provision new databases autonomously using this exact key.

typescript
import { Pulsbase } from '@pulsbase/sdk';

const client = new Pulsbase({
  apiKey: process.env.PULSBASE_API_KEY,
});

// Autonomous AI Agent Cluster Provisioning
const cluster = await client.databases.create({
  name: 'production-ai-lake',
  jurisdiction: 'global', // 'global' | 'eu' | 'us-east' | 'us-west' | 'apac'
});
3

Insert & Embed

When you insert text, Pulsbase automatically generates high-dimensional embeddings in the background using state-of-the-art embedding models.

typescript
await db.sql.execute(
  'db_123',
  'INSERT INTO documents (id, content, metadata) VALUES (?, ?, ?)',
  ['doc_1', 'Pulsbase handles semantic embeddings automatically at the edge.', JSON.stringify({ category: 'docs' })]
);
4

Hybrid Search

Perform blazingly fast searches that combine vector similarity with traditional keyword filtering in a single query.

typescript
const results = await db.search.query({
  database_id: 'db_123',
  query: 'How does embedding work?',
  limit: 5
});

console.log(results);

Ready for more?

Explore the full API reference or integrate with AI agents.

API Reference SDK Guides