n8n Workflows

Notes on building automation with n8n. I am a certified n8n professional and use it for everything from lead capture to AI content pipelines.

What n8n Is Good For

• Connecting APIs without writing glue code
• Automating repetitive manual tasks
• Building AI pipelines that process data through multiple steps
• Webhook receivers that trigger on external events
• Scheduled jobs (daily reports, weekly digests, data syncs)

Not great for: real-time latency-sensitive operations, complex stateful workflows, or anything needing sub-second response times.

Core Concepts

Lead Capture Workflow

Webhook receives a form submission, saves to database, sends welcome email, notifies Slack.

Webhook trigger
  → Set node (clean and normalize fields)
  → Airtable node (save lead)
  → Gmail node (send welcome email)
  → Slack node (notify #leads channel)

// Webhook payload example
{
  "name": "Wayan Phantom",
  "email": "wayanphantomme@gmail.com",
  "message": "Interested in AI automation"
}

In the Gmail node, use expressions to personalize:

Subject: Welcome, {{ $json.name }}
Body: Hi {{ $json.name }}, thanks for reaching out...

AI Content Pipeline (Impact Stories)

Automated blog platform that generates and publishes articles daily.

Schedule trigger (every day at 8am)
  → HTTP Request (fetch topic/source from CMS)
  → OpenAI node (generate article draft)
  → Code node (format, add metadata, sanitize)
  → HTTP Request POST (publish to CMS API)
  → Telegram node (notify me it published)

The Code node does the heavy lifting — stripping unsafe HTML, generating slugs, adding frontmatter:

// Code node (JavaScript)
const title = $input.item.json.title;
const content = $input.item.json.content;

return [{
  json: {
    slug: title.toLowerCase().replace(/\s+/g, "-"),
    publishedAt: new Date().toISOString(),
    content: content.trim(),
    wordCount: content.split(" ").length,
  }
}];

HTTP Request Node

The most used node. Calls any REST API.

Method: POST
URL: https://api.example.com/articles
Authentication: Header Auth
  Header Name: Authorization
  Header Value: Bearer {{ $credentials.apiKey }}
Body (JSON):
  {
    "title": "{{ $json.title }}",
    "content": "{{ $json.content }}"
  }

Webhook + AI Agent Pattern

Useful for AI-powered Slack bots, customer support, or any chat-based workflow.

Webhook trigger (POST from Slack/Telegram)
  → Code node (extract user message)
  → HTTP Request (POST /api/chat with message history)
  → Code node (format response)
  → HTTP Request (POST reply back to Slack/Telegram)

The /api/chat endpoint is your own Next.js route handler with the LLM logic.

Tips

• Use Split In Batches node when processing arrays — don't try to process 1000 items in one shot
• Store credentials in n8n credential store, not hardcoded in nodes
• Use Error Workflow to catch failures and send yourself a Telegram alert
• Self-hosted n8n on Railway or Render is cheap and gives you full control

Last updated: September 2026.