Workflow Automation Tools

// n8n workflow structure (simplified)
// Import this JSON into n8n to recreate the workflow
{
  "name": "AI Video Pipeline",
  "nodes": [
    {
      "name": "Schedule Trigger",
      "type": "n8n-nodes-base.scheduleTrigger",
      "parameters": {
        "rule": { "interval": [{ "field": "hours", "hoursInterval": 6 }] }
      }
    },
    {
      "name": "Get Topic from Sheet",
      "type": "n8n-nodes-base.googleSheets",
      "parameters": {
        "operation": "read",
        "sheetId": "YOUR_SHEET_ID",
        "range": "Calendar!A:E",
        "filters": { "status": "pending" }
      }
    },
    {
      "name": "Generate Script (GPT-4)",
      "type": "@n8n/n8n-nodes-langchain.openAi",
      "parameters": {
        "model": "gpt-4o",
        "prompt": "Write a 60-second video script about: {{ $json.topic }}"
      }
    },
    {
      "name": "Generate Images (DALL-E)",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "method": "POST",
        "url": "https://api.openai.com/v1/images/generations",
        "body": { "prompt": "{{ $json.visual_prompt }}", "model": "dall-e-3" }
      }
    },
    {
      "name": "Generate Voice (ElevenLabs)",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "method": "POST",
        "url": "https://api.elevenlabs.io/v1/text-to-speech/pNInz6obpgDQGcFmaJgB"
      }
    },
    {
      "name": "Generate Video Clips (Runway)",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "method": "POST",
        "url": "https://api.runwayml.com/v1/image_to_video"
      }
    },
    {
      "name": "Assemble with FFmpeg",
      "type": "n8n-nodes-base.executeCommand",
      "parameters": {
        "command": "ffmpeg -i clips.txt -i voice.mp3 -c:v copy -c:a aac output.mp4"
      }
    },
    {
      "name": "Upload to YouTube",
      "type": "n8n-nodes-base.youTube",
      "parameters": {
        "operation": "upload",
        "title": "{{ $json.title }}",
        "description": "{{ $json.description }}"
      }
    },
    {
      "name": "Notify on Slack",
      "type": "n8n-nodes-base.slack",
      "parameters": {
        "channel": "#video-pipeline",
        "text": "New video published: {{ $json.youtube_url }}"
      }
    }
  ]
}

// Make Scenario Structure:
//
// [Schedule] → [Google Sheets: Get Row] → [OpenAI: Chat Completion]
//     ↓
// [JSON: Parse Script] → [Iterator: Loop Scenes]
//     ↓
// [OpenAI: Create Image] → [HTTP: Runway Image-to-Video]
//     ↓
// [Sleep: Wait 60s] → [HTTP: Runway Check Status] → [Router]
//     ↓                                                ↓
// [If complete: Download Video]              [If processing: Loop back]
//     ↓
// [Aggregator: Collect All Clips]
//     ↓
// [HTTP: FFmpeg Assembly API] → [Google Drive: Upload]
//     ↓
// [YouTube: Upload Video] → [Slack: Send Notification]

// Make Module Config Example (OpenAI Chat Completion):
{
  "module": "openai.createChatCompletion",
  "parameters": {
    "model": "gpt-4o",
    "messages": [
      {
        "role": "system",
        "content": "Write a structured video script with scenes..."
      },
      {
        "role": "user",
        "content": "Topic: {{1.topic}}"
      }
    ],
    "response_format": "json_object",
    "max_tokens": 2000
  }
}

// Zapier Zap Configuration:
//
// TRIGGER: New Video Published on YouTube
//   App: YouTube
//   Event: New Video in Channel
//
// ACTION 1: Generate Social Post with ChatGPT
//   App: ChatGPT (by Zapier)
//   Event: Conversation
//   Prompt: "Write a Twitter post (max 280 chars) promoting this video:
//            Title: {{trigger.title}}
//            Description: {{trigger.description}}
//            URL: {{trigger.url}}"
//
// ACTION 2: Post to X (Twitter)
//   App: Twitter
//   Event: Create Tweet
//   Text: {{chatgpt.reply}} + {{trigger.url}}
//
// ACTION 3: Post to LinkedIn
//   App: LinkedIn
//   Event: Create Share Update
//   Content: "New video: {{trigger.title}}\n\n{{trigger.description}}\n\n{{trigger.url}}"
//
// ACTION 4: Send Slack Notification
//   App: Slack
//   Event: Send Channel Message
//   Channel: #content-published
//   Message: "Published & distributed: {{trigger.title}} — {{trigger.url}}"

// Daily Video Pipeline — n8n Workflow
// Trigger: Every day at 6:00 AM UTC

1. [Schedule Trigger] → fires at 06:00 UTC daily

2. [Google Sheets] → Read first 'pending' row from Calendar sheet
   Output: { topic, keywords, platform, publishTime }

3. [OpenAI GPT-4o] → Generate structured script
   Input: topic + keywords
   Output: { title, scenes: [...], total_duration }

4. [Split In Batches] → Loop over scenes array
   For each scene:
   ├─ [OpenAI DALL-E 3] → Generate scene image
   ├─ [HTTP: Runway] → Submit image-to-video task
   ├─ [Wait 90s] → Pause for Runway processing
   └─ [HTTP: Runway Status] → Check & download completed video

5. [ElevenLabs] → Generate full voiceover (parallel with step 4)
   Input: All narration text concatenated
   Output: voiceover.mp3

6. [Merge] → Combine all video clips + voiceover path

7. [Execute Command] → FFmpeg assembly
   ffmpeg -f concat -i clips.txt -i voiceover.mp3 \
     -c:v libx264 -c:a aac final.mp4

8. [YouTube Upload] → Upload final.mp4 with metadata
   Schedule: publishTime from calendar

9. [Google Sheets] → Update row status to 'published'

10. [Slack] → Send notification: "Video published: {title}"

// n8n Error Trigger Workflow
// This separate workflow fires whenever ANY workflow errors

1. [Error Trigger] → Catches errors from all workflows
   Output: { workflow_name, execution_id, error_message, node_name }

2. [Slack] → Post to #pipeline-errors channel
   Message format:
   "🔴 Pipeline Error
   Workflow: {{ $json.workflow.name }}
   Node: {{ $json.execution.lastNodeExecuted }}
   Error: {{ $json.execution.error.message }}
   Execution: https://n8n.your-server.com/execution/{{ $json.execution.id }}"

3. [Google Sheets] → Log error to Error Tracking sheet
   Columns: timestamp, workflow, node, error, execution_url, resolved(false)

# Install n8n with Docker (takes 30 seconds)
docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -v n8n_data:/home/node/.n8n \
  n8nio/n8n

# Open http://localhost:5678 in your browser
# Create your first workflow:
# 1. Click '+' to add a Schedule Trigger node
# 2. Add an HTTP Request node (connect to OpenAI API)
# 3. Add another HTTP Request node (connect to DALL-E API)
# 4. Add a Code node for FFmpeg processing
# 5. Add a YouTube node for publishing
# 6. Connect the nodes and activate the workflow

# For production, use Docker Compose with PostgreSQL:
# docker-compose.yml with n8n + postgres + nginx reverse proxy