Skip to content

Quickstart

Dmitrii Karataev edited this page Feb 26, 2026 · 3 revisions

Quickstart

Get RocketRide running and connected in under five minutes.

1. Prerequisites

You need Node.js 18+ installed. Everything else (pnpm, CMake, Python, JDK, Tika) is downloaded automatically by the build system.

Verify Node.js:

node --version   # Must be v18.0.0 or higher

2. Clone and Build

git clone https://github.com/rocketride-org/rocketride-server.git
cd rocketride-server
./builder build

The first build downloads all toolchains and dependencies. Subsequent builds are incremental and much faster. See Installation for platform-specific notes.

3. Verify the Engine

./Engine --verify

A successful verification confirms the engine binary, pipeline nodes, and all dependencies are correctly installed.

4. Run the Engine as a Service

./Engine --service --port 5565

The engine is now listening for WebSocket connections on port 5565.

5. Connect with the TypeScript SDK

npm install rocketride
import { RocketRideClient, Question } from 'rocketride';

const client = new RocketRideClient({
  uri: 'http://localhost:5565',
  auth: 'your-api-key',
});

await client.connect();

// Load and run a pipeline
const result = await client.use({ filepath: 'pipeline.json' });

// Send data for processing
await client.send(result.token, 'Hello, RocketRide!');

// Chat with AI
const question = new Question();
question.addQuestion('What did I just send?');
const answer = await client.chat({
  token: result.token,
  question: question,
});

await client.disconnect();

6. Connect with the Python SDK

pip install rocketride
from rocketride import RocketRideClient, Question

async with RocketRideClient(uri='http://localhost:5565', auth='your-api-key') as client:
    # Load and run a pipeline
    result = await client.use(filepath='pipeline.json')

    # Send data for processing
    await client.send(result['token'], 'Hello, RocketRide!')

    # Chat with AI
    question = Question()
    question.addQuestion('What did I just send?')
    answer = await client.chat(token=result['token'], question=question)

Next Steps

Clone this wiki locally