Getting Started

What is NodeGraph?

NodeGraph is a visual, node-based platform for building AI-powered automation workflows. You construct workflows by placing blocks on a canvas and connecting them with edges (routes). The system handles the rest: voice calls, trading, messaging, monitoring, and AI processing.

Product URL: app.nodegraph.io

Core Concepts

Workspace

A workspace is your project. It contains a canvas with blocks and edges. Each user can have multiple workspaces.

Blocks

Blocks are the building units. Each block has a type and a purpose:

• Voice & Telephony -- softphone, live voice agents (Gemini Live, OpenAI Realtime), call guard, translator, trunks
• AI Agents -- LLM hubs (Claude, OpenAI, Grok, DeepSeek, Groq) and personas
• Messengers -- Telegram, Discord, WhatsApp, Matrix
• Automation -- Workers (Python code), Schedulers, Logic Gates, Webhooks
• Data & Markets -- market feeds, web scraping, and exchange connectors
• UI / Monitoring -- live dashboards for calls, workers, and sessions

See the full list in the Block Catalog.

Edges (Routes)

Edges connect blocks together. This is the core principle:

Edge = Connection. When you draw an edge from Block A to Block B, they become linked. Delete the edge -- they disconnect. No manual config needed.

For example:

• Draw edge from Worker to Bybit block -- ctx.bybit becomes available in your worker code
• Draw edge from Worker to Monitor block -- ctx.monitor lets you render dashboards
• Draw edge from Worker to Telegram block -- ctx.telegram lets you send messages

Workers

Workers are Python scripts that run continuously (24/7) in a sandboxed environment. They follow a simple tick-based model:

def setup(ctx):
    """Called once when the worker starts."""
    ctx.log.info("Starting up!")

def tick(ctx):
    """Called every N seconds (configurable)."""
    price = ctx.bybit.get_tickers("BTCUSDT")
    ctx.monitor.render([
        ctx.monitor.metric("BTC", price["last_price"])
    ])

Learn more in the Workers section.

Quick Start

1. Open app.nodegraph.io and create an account
2. You'll land on your workspace canvas
3. Click + Block to add blocks
4. Click Route to connect blocks with edges
5. Open a Worker block, write Python code, and click Deploy
6. Click Start to run your worker

That's it! Your code runs in the cloud, 24/7.