In this hands-on tutorial, you'll build an automated website content analyzer that discovers all pages on a website, extracts their content, and generates AI-powered insights. This is a practical example of combining multiple nodes to create a powerful workflow.

What You'll Build

By the end of this tutorial, you'll have a flow that:

• Maps all pages on a website
• Scrapes content from each page
• Cleans HTML into readable text
• Analyzes each page with AI
• Generates a comprehensive site summary
• Saves all results to a dataset

Prerequisites

• Evaligo account (free tier works fine)
• Basic familiarity with the platform interface
• A prompt already created (we'll guide you)

Step 1: Create Your Prompts

First, we need to create the prompts we'll use in our flow.

Prompt 1: Page Analyzer

1. 1

   Go to Playground Navigate to the Prompt Playground

2. 2

   Create new prompt Click "New Prompt"

3. 3

   Enter prompt template Copy the template below

4. 4

   Test it Try with sample content

5. 5

   Save as "Page Analyzer" Save for use in flow

Analyze this webpage content and provide a structured summary:

URL: {{url}}
Content: {{content}}

Provide:
1. Main topic or purpose
2. Key points (3-5 bullet points)
3. Target audience
4. Content type (article, product page, documentation, etc.)

Be concise and factual.

Prompt 2: Site Summarizer

Create a second prompt for summarizing all pages:

Analyze these page summaries from a website and create a comprehensive overview:

{{pageAnalyses}}

Provide:
1. Overall purpose of the website
2. Main content categories
3. Target audience
4. Key themes and topics
5. Recommendations for content improvement

Total pages analyzed: {{pageCount}}

Save this as "Site Summarizer".

Step 2: Build the Flow

Now let's build the actual workflow.

Add API Input Node

1. 1

   Open Flow Playground Navigate to AI Flows

2. 2

   Drag API Input node From the node palette

3. 3

   Configure input Add field "websiteUrl" (string, required)

Add Website Mapper Node

1. 1

   Drag Website Mapper node Add to canvas

2. 2

   Connect API Input to Website Mapper Drag from API Input's output handle

3. 3

   Map variable Map _input.websiteUrl to url

4. 4

   Configure settings Set max pages: 10 (for testing)

Add Array Splitter

1. 1

   Drag Array Splitter node This will process each URL individually

2. 2

   Connect Website Mapper output Connect to Array Splitter input

3. 3

   Map array Map out.urls to the splitter

Add Page Scraper

1. 1

   Drag Page Scraper node Add after Array Splitter

2. 2

   Connect nodes Each split URL will be scraped

3. 3

   Map URL Map out (the URL) to url

Add HTML Text Extractor

1. 1

   Drag HTML Text Extractor Clean the scraped HTML

2. 2

   Connect Page Scraper Connect output to extractor

3. 3

   Map HTML Map out.html to html

Add Prompt Node (Page Analyzer)

1. 1

   Drag Prompt node Add to canvas

2. 2

   Select "Page Analyzer" Choose from saved prompts

3. 3

   Map variables Map out.text to content
   Map original URL to url

Add Array Flatten

1. 1

   Drag Array Flatten node Collect all page analyses

2. 2

   Connect Prompt output Automatically collects all results

Add Second Prompt (Site Summarizer)

1. 1

   Drag another Prompt node Add below Array Flatten

2. 2

   Select "Site Summarizer" Choose from saved prompts

3. 3

   Map flattened results Map out (array) to pageAnalyses
   Map out.count to pageCount

Add Dataset Sink

1. 1

   Drag Dataset Sink Save results

2. 2

   Connect final prompt Connect Site Summarizer output

3. 3

   Create new dataset "Website Analyses"

4. 4

   Map fields Map summary and metadata to dataset columns

Add API Output

1. 1

   Drag API Output node Define what the API returns

2. 2

   Map output fields summary: from Site Summarizer
   pagesAnalyzed: from Array Flatten count
   websiteUrl: from original input

Step 3: Test Your Flow

Run in Playground

1. 1

   Click "Run Flow" Button in top toolbar

2. 2

   Enter test URL Try "https://evaligo.com" or your own site

3. 3

   Watch execution See nodes light up as they process

4. 4

   Review results Click nodes to inspect outputs

Verify Output

Check that you receive:

• Individual page analyses in Array Flatten
• Comprehensive summary in Site Summarizer
• Data saved in Dataset Sink
• Structured JSON in API Output

Step 4: Deploy as API

1. 1

   Click "Deploy" Button in top toolbar

2. 2

   Name your API "Website Content Analyzer"

3. 3

   Copy API endpoint You'll use this to call your flow

4. 4

   Copy API key For authentication

Test API Call

curl -X POST https://api.evaligo.com/flows/your-flow-id/execute-async \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "websiteUrl": "https://example.com"
  }'

Step 5: Monitor and Improve

Review Results

• Check the "Website Analyses" dataset
• Review individual page analyses for quality
• Check site summary for accuracy
• Note execution time and costs

Optimize Performance

Consider these improvements:

• Enable parallel processing for Array Splitter
• Add caching for frequently analyzed sites
• Filter out certain page types (login, search)
• Increase max pages for full site analysis

Enhance Analysis

• Add sentiment analysis
• Extract structured data (prices, dates)
• Compare changes over time
• Generate SEO recommendations

Common Issues

Empty Results

Problem: No pages found or scraped

Solution:

• Verify URL is accessible
• Check Website Mapper output
• Try a different website
• Increase timeout settings

Timeout Errors

Problem: Flow times out

Solution:

• Reduce max pages temporarily
• Use async execution mode
• Enable parallel processing

Poor Analysis Quality

Problem: AI summaries are generic

Solution:

• Refine prompts with more specific instructions
• Test prompts in Playground first
• Add examples to prompt template
• Use structured output schema

Next Steps

Extend This Flow

• Add competitor analysis (compare multiple sites)
• Schedule regular scans to detect changes
• Generate reports in different formats
• Build a dashboard with historical data

Learn More

• Try the SEO Audit Tutorial
• Explore batch processing patterns
• Set up monitoring and alerts
• Integrate with your existing tools