Back to .md Directory

Automated Health Content Generation System

**Problem**: Creating evidence-based health content is boring and time-consuming

May 2, 2026
0 downloads
2 views
ai prompt claude workflow automation
View source

Automated Health Content Generation System

Stop Manual Content Creation - Let AI + n8n Do the Work

Problem: Creating evidence-based health content is boring and time-consuming Solution: Automated pipeline that scrapes campaigns → verifies sources → generates posts → creates images → schedules


🤖 AUTOMATION ARCHITECTURE

┌─────────────────────┐
│ Campaign Websites   │ (Movember, Prostate Cancer UK, THT, etc.)
└──────────┬──────────┘
           │ n8n scrapes verified stats
           ▼
┌─────────────────────┐
│ Source Database     │ (Airtable/Supabase: campaign quotes + sources)
└──────────┬──────────┘
           │ AI formats with "Move this Movember" framing
           ▼
┌─────────────────────┐
│ Claude/GPT-4        │ (Generates posts with verified citations)
└──────────┬──────────┘
           │ DALL-E/Midjourney generates graphics
           ▼
┌─────────────────────┐
│ Image Generation    │ (BLKOUT brand colors + quote overlays)
└──────────┬──────────┘
           │ Late.dev schedules posts
           ▼
┌─────────────────────┐
│ Social Media        │ (Instagram, Twitter, LinkedIn, Facebook)
└─────────────────────┘

🔧 IMPLEMENTATION OPTIONS

Option 1: Full n8n Automation Workflow (RECOMMENDED)

Best for: Completely hands-off content generation

n8n Workflow Steps:

  1. Campaign Scraper (Weekly)

    • Scrape Movember UK statistics page
    • Scrape Prostate Cancer UK Black men's page
    • Scrape THT PrEP statistics
    • Scrape 56 Dean Street service data
    • Store in Supabase with source URLs + dates
  2. Content Generator (On-demand/Scheduled)

    • Read campaign stats from database
    • Send to Claude API with prompt: "Create 'Move this Movember' post using this statistic: [stat]. Include source citation. Target: Black queer men."
    • Validate output has proper citation
    • Store generated post in content library
  3. Image Generator (Automated)

    • Send post text to DALL-E/Midjourney
    • Prompt: "Create social media graphic with BLKOUT brand (Liberation Teal #008B8B, Community Gold #FFD700, Black background). Include quote: [text]. Medical imagery, Black queer men, professional NHS style."
    • Add text overlay with quote + source
    • Save to asset library
  4. Scheduler & Poster (Daily 9am)

    • Pull next post from queue
    • Post to Instagram, Twitter, LinkedIn via APIs
    • Log posted content
    • Send confirmation notification

Time Saved: 30 hours/month → 2 hours/month (setup + monitoring)


Option 2: IVOR-Powered Content Assistant

Best for: On-demand content creation with IVOR backend

Flow:

  1. User asks IVOR: "Create a Movember post about prostate cancer for Black men"
  2. IVOR:
    • Queries campaign database for verified Prostate Cancer UK statistics
    • Formats post with "Move this Movember" framing
    • Adds proper citations
    • Generates image via API
    • Returns ready-to-post content

Implementation:

# IVOR Health Content Module
def generate_health_post(topic, target_audience):
    # Query verified stats database
    stats = query_campaign_database(topic)

    # Generate post with Claude
    post = claude_api.generate(
        prompt=f"Create 'Move this Movember' post about {topic} for {target_audience}. Use this verified statistic: {stats['quote']}. Source: {stats['source']}"
    )

    # Generate image
    image = dalle_api.generate(
        prompt=f"BLKOUT brand social media graphic. {topic}. Black queer men. Medical professional style. Teal and gold colors. Quote overlay: {post}"
    )

    return {
        "post": post,
        "image": image,
        "source": stats['source_url'],
        "organization": stats['campaign']
    }

Time Saved: Create posts in 30 seconds instead of 30 minutes


Option 3: Hybrid Semi-Automated System

Best for: You review before posting, but AI does the heavy lifting

Tools:

  • Airtable: Campaign stats database (manual entry once/quarter)
  • Make.com/Zapier: Trigger content generation on schedule
  • Claude API: Generate posts from database
  • Canva API/Bannerbear: Auto-generate graphics
  • Late.dev: Schedule & review before posting

Workflow:

  1. Quarterly: Manually add 10-15 verified campaign stats to Airtable
  2. Weekly: Trigger content generation (creates 7 posts automatically)
  3. Daily: Review & approve posts in Late.dev dashboard
  4. Automated: Posts go live at scheduled times

Time Saved: 20 hours/month → 4 hours/month


📦 READY-TO-USE N8N WORKFLOW

I can create a complete n8n workflow that:

Input (Airtable/Google Sheets):

Topic,Campaign,Quote,Source URL,Target Date
Prostate Cancer,Prostate Cancer UK,"1 in 4 Black men will be diagnosed",https://prostatecanceruk.org,2025-11-02
HIV Prevention,NHS England,"PrEP reduced HIV risk by 86%",https://gov.uk/prep,2025-11-08

n8n Processing:

  1. Read row from spreadsheet
  2. Generate post: "Move this Movember - [topic]. [Campaign] says: '[quote]'. [Call to action]. Source: [URL]"
  3. Generate image with quote overlay
  4. Schedule to Late.dev for target date
  5. Mark row as "processed"

Output:

  • 30 posts generated in 5 minutes
  • Images ready
  • Scheduled across November
  • All sources cited

🎨 AUTOMATED IMAGE GENERATION

Option A: DALL-E with Consistent Branding

# DALL-E prompt template
def generate_health_graphic(quote, topic):
    prompt = f"""
    Professional medical social media graphic for LGBTQ+ health campaign.

    Style: Modern, NHS-inspired, empowering
    Colors: Liberation Teal (#008B8B), Community Gold (#FFD700), Black background
    Subject: {topic} - Black queer men in healthcare setting
    Text overlay: "{quote}"
    Typography: Bold, clear, medical-style fonts
    Mood: Destigmatizing, professional, affirming
    Layout: Clean, accessible, high contrast
    """
    return dalle_api.generate(prompt)

Option B: Canva API with Templates

  1. Create 5-10 Canva templates with BLKOUT branding
  2. Use Canva API to auto-populate with different quotes
  3. Export as PNG for social media
  4. Consistent branding, zero design work

Option C: Bannerbear (Fastest)

  • Set up templates with dynamic text zones
  • API auto-generates images in seconds
  • Perfect for high-volume content

Time Saved: 10 hours/month on design → 0 hours (fully automated)


🚀 QUICK START: 30-MINUTE SETUP

Minimum Viable Automation:

  1. Create Google Sheet (5 min)

    • Columns: Topic | Campaign | Quote | Source | Date | Status
    • Add your 30 verified stats
  2. Set up n8n Workflow (15 min)

    • Trigger: Schedule (daily 8am)
    • Read next unprocessed row from Google Sheet
    • HTTP Request to Claude API with template prompt
    • HTTP Request to DALL-E with image prompt
    • Save outputs to folder
    • Mark row as "processed"
  3. Connect Late.dev (10 min)

    • Upload generated content
    • Schedule posts
    • Enable auto-posting

Done! Your November content is automated.


💰 COST BREAKDOWN

Monthly Running Costs:

  • Claude API: ~£10 (30 posts × 500 tokens)
  • DALL-E: ~£15 (30 images)
  • n8n Cloud: £0 (free tier) or £20 (pro)
  • Late.dev: £0-25 (depends on plan)
  • Total: £25-50/month

ROI: You save 20-30 hours/month Value: £500-1,500/month (at £25-50/hour)


🔐 MAINTAINING ACCURACY WITH AUTOMATION

Quality Control Checkpoints:

  1. Source Verification Layer

    • n8n checks source URLs are still active
    • Flags broken links for manual review
    • Validates stat hasn't changed on source page
  2. Citation Requirement

    • Claude prompt MUST include source attribution
    • Workflow rejects outputs without citations
    • All posts include "Source: [Campaign Name] [Year]"
  3. Human Review Gate (Optional)

    • Store generated posts in "Review" queue
    • Slack/Email notification: "5 posts ready for review"
    • Approve with one click → auto-posts
    • Reject → regenerates with feedback
  4. Quarterly Audit

    • Review campaign websites for updated stats
    • Update source database
    • Regenerate any outdated content

🎯 RECOMMENDED APPROACH FOR BLKOUT

Phase 1 (Week 1): Foundation

  • Set up Airtable with 50 verified campaign stats
  • Create Claude API integration
  • Build basic n8n content generator

Phase 2 (Week 2): Automation

  • Add DALL-E image generation
  • Connect to Late.dev scheduler
  • Test end-to-end workflow

Phase 3 (Week 3): IVOR Integration

  • Build IVOR module: /health-post [topic]
  • Users can request health content on-demand
  • IVOR generates using campaign database

Phase 4 (Week 4): Scale

  • Expand to other health months (Breast Cancer Awareness, World AIDS Day, etc.)
  • Add multi-language support
  • Create API for other organizations to use

📋 NEXT STEPS

Choose your automation level:

  • Full Automation: I'll build the complete n8n workflow (1-2 hours)
  • IVOR Integration: Add health content generation to IVOR (2-3 hours)
  • Hybrid: Set up the database + templates, you review & approve (30 min)

What you need to provide:

  • n8n instance access or credentials
  • Claude API key
  • DALL-E/image generation API key
  • Late.dev or social media API credentials
  • Google Sheet/Airtable for campaign stats database

I can build this in one session if you want to proceed.


🎁 BONUS: REUSABLE TEMPLATE

Once built, this system works for:

  • ✅ Black History Month posts
  • ✅ Pride Month content
  • ✅ World AIDS Day messaging
  • ✅ Mental Health Awareness Week
  • ✅ Any health campaign with verified sources

One-time setup → Infinite automated content 🚀


Stop wasting time on manual content creation. Let's automate this right now.

Ready to build?

Related Documents