Search Company Information with Carnot Search in Jinba Flow

Use CARNOT_SEARCH_CORP and CARNOT_SEARCH_CORP_PRO to search for company information in a Jinba Flow workflow.

Carnot Search provides powerful company search capabilities to find detailed business information. With both standard and Pro versions, you can access comprehensive company databases for market research, lead generation, and business intelligence.

Tools

CARNOT_SEARCH_CORP

• Basic company search functionality
• Find company information by name or criteria
• Retrieve company URLs and addresses
• Configurable result count
• Inputs: query (search text), num (number of results, default 10) — Output: result.items, an array of { name, url, address }

CARNOT_SEARCH_CORP_PRO

• Triggers a Carnot Search Corp Pro job on Carnot’s external server; the job itself runs outside the flow
• Inputs: spreadsheets_key (required; the key of the target spreadsheet), and optional mail_runner_flow_url, mail_runner_flow_secret, and slack_channel_id
• Output: result.content is Job triggered successfully or Failed to trigger job

Authentication

CARNOT_SEARCH_CORP requires API keys from both Perplexity and OpenAI (perplexity_api_key and openai_api_key config). CARNOT_SEARCH_CORP_PRO requires a Carnot API key (carnot_api_key config).

Required credentials:

• Perplexity API Key and OpenAI API Key (CARNOT_SEARCH_CORP)
• Carnot API Key (CARNOT_SEARCH_CORP_PRO)

Note: Treat API keys as sensitive information and never commit them to public repositories.

Run a basic company search

- id: search_companies
  name: search_companies
  tool: CARNOT_SEARCH_CORP
  config:
    - name: perplexity_api_key
      value: "{{secrets.PERPLEXITY_API_KEY}}"
    - name: openai_api_key
      value: "{{secrets.OPENAI_API_KEY}}"
  input:
    - name: query
      value: "software companies in San Francisco"
    - name: num
      value: 10

- id: process_search_results
  name: process_search_results
  tool: PYTHON_SANDBOX_RUN
  needs: [search_companies]
  input:
    - name: code
      value: |
        import json
        
        # Process search results
        results = steps.search_companies.result["items"]
        
        print("Company Search Results:")
        print("=" * 50)
        
        for i, company in enumerate(results, 1):
            print(f"{i}. {company.get('name', 'Unknown')}")
            print(f"   URL: {company.get('url', 'N/A')}")
            print(f"   Address: {company.get('address', 'N/A')}")
            print()
        
        print(f"Total companies found: {len(results)}")

Trigger a Pro job

- id: trigger_pro_job
  name: trigger_pro_job
  tool: CARNOT_SEARCH_CORP_PRO
  config:
    - name: carnot_api_key
      value: "{{secrets.CARNOT_API_KEY}}"
  input:
    - name: spreadsheets_key
      value: "1AbCdEfGhIjKlMnOpQrStUvWxYz0123456789"
    - name: slack_channel_id
      value: "C0123456789"

# steps.trigger_pro_job.result.content is "Job triggered successfully" or "Failed to trigger job"

Analyze and export search results

- id: advanced_company_search
  name: advanced_company_search
  tool: CARNOT_SEARCH_CORP
  config:
    - name: perplexity_api_key
      value: "{{secrets.PERPLEXITY_API_KEY}}"
    - name: openai_api_key
      value: "{{secrets.OPENAI_API_KEY}}"
  input:
    - name: query
      value: "fintech startups with Series A funding in New York"
    - name: num
      value: 20

- id: analyze_competitive_landscape
  name: analyze_competitive_landscape
  tool: OPENAI_INVOKE
  needs: [advanced_company_search]
  config:
    - name: version
      value: gpt-5.2
  input:
    - name: prompt
      value: |
        Analyze the following company search results and provide insights:
        
        {{steps.advanced_company_search.result.items | json}}
        
        Please provide:
        1. Market trends analysis
        2. Geographic distribution
        3. Company size patterns
        4. Key players identification
        5. Competitive landscape overview

- id: export_company_data
  name: export_company_data
  tool: CONVERTER_ARRAY_TO_CSV
  needs: [advanced_company_search]
  input:
    - name: input
      value: |
        [
          ["Company Name", "Website", "Address", "Search Score"],
          {% for company in steps.advanced_company_search.result.items %}
          ["{{company.name}}", "{{company.url}}", "{{company.address}}", "High"]{% if not loop.last %},{% endif %}
          {% endfor %}
        ]

Example: Market Research Workflow

- id: search_competitors
  name: search_competitors
  tool: CARNOT_SEARCH_CORP
  config:
    - name: perplexity_api_key
      value: "{{secrets.PERPLEXITY_API_KEY}}"
    - name: openai_api_key
      value: "{{secrets.OPENAI_API_KEY}}"
  input:
    - name: query
      value: "{{input.industry}} companies in {{input.target_region}}"
    - name: num
      value: 50

- id: enrich_company_data
  name: enrich_company_data
  tool: PYTHON_SANDBOX_RUN
  needs: [search_competitors]
  input:
    - name: code
      value: |
        import json
        import re
        from urllib.parse import urlparse
        
        companies = json.loads(r'''{{steps.search_competitors.result.items}}''')
        enriched_data = []
        
        for company in companies:
            # Extract domain from URL
            domain = ""
            if company.get('url'):
                try:
                    domain = urlparse(company['url']).netloc
                except:
                    domain = company['url']
            
            # Analyze company size based on address/description
            size_indicator = "Unknown"
            address = company.get('address', '').lower()
            if any(term in address for term in ['suite', 'floor', 'building']):
                size_indicator = "Medium-Large"
            elif any(term in address for term in ['street', 'ave', 'road']):
                size_indicator = "Small-Medium"
            
            enriched = {
                "name": company.get('name', ''),
                "url": company.get('url', ''),
                "domain": domain,
                "address": company.get('address', ''),
                "estimated_size": size_indicator
            }
            enriched_data.append(enriched)
        
        print(json.dumps(enriched_data, indent=2))

- id: create_market_report
  name: create_market_report
  tool: OPENAI_INVOKE
  needs: [enrich_company_data]
  config:
    - name: version
      value: gpt-5.2
  input:
    - name: prompt
      value: |
        Create a comprehensive market research report based on this company data:
        
        {{steps.enrich_company_data.result.data}}
        
        Include:
        1. Executive Summary
        2. Market Size Assessment
        3. Key Players Analysis
        4. Geographic Distribution
        5. Competitive Landscape
        6. Market Opportunities
        7. Recommendations
        
        Format as a professional business report.

- id: save_market_report
  name: save_market_report
  tool: OUTPUT_FILE
  needs: [create_market_report]
  input:
    - name: content
      value: "{{steps.create_market_report.result.content}}"
    - name: filename
      value: "market_research_report.txt"
    - name: fileType
      value: "txt"

Build a lead generation pipeline

- id: find_potential_leads
  name: find_potential_leads
  tool: CARNOT_SEARCH_CORP
  config:
    - name: perplexity_api_key
      value: "{{secrets.PERPLEXITY_API_KEY}}"
    - name: openai_api_key
      value: "{{secrets.OPENAI_API_KEY}}"
  input:
    - name: query
      value: "{{input.target_industry}} companies with 10-100 employees"
    - name: num
      value: 25

- id: qualify_leads
  name: qualify_leads
  tool: OPENAI_INVOKE
  needs: [find_potential_leads]
  config:
    - name: version
      value: gpt-5.2
  input:
    - name: prompt
      value: |
        Analyze these potential leads and score them for sales qualification:
        
        {{steps.find_potential_leads.result.items | json}}
        
        Scoring criteria:
        - Company size and growth potential (1-10)
        - Industry fit (1-10)
        - Geographic accessibility (1-10)
        - Technology alignment (1-10)
        
        Return a scored list with recommendations for follow-up.

- id: export_qualified_leads
  name: export_qualified_leads
  tool: CONVERTER_ARRAY_TO_CSV
  needs: [find_potential_leads]
  input:
    - name: input
      value: |
        [
          ["Company", "Website", "Address", "Qualification Score", "Priority", "Notes"],
          {% for company in steps.find_potential_leads.result.items %}
          ["{{company.name}}", "{{company.url}}", "{{company.address}}", "TBD", "Medium", "Follow up required"]{% if not loop.last %},{% endif %}
          {% endfor %}
        ]

Search Parameters

Query Optimization

• Use specific industry terms
• Include geographic qualifiers
• Specify company size when relevant
• Use boolean operators for complex searches

Result Management

• Start with smaller result sets (10-20)
• Increase for comprehensive research (50+)
• Consider API rate limits
• Balance accuracy vs coverage

Use Cases

• Market Research: Identify competitors and market players
• Lead Generation: Find potential customers and partners
• Investment Research: Analyze market sectors and opportunities
• Competitive Intelligence: Monitor competitor landscape
• Partnership Development: Discover potential business partners
• Sales Prospecting: Build targeted prospect lists
• Due Diligence: Research companies for business decisions

Data Structure

CARNOT_SEARCH_CORP returns result.items; each item includes:

{
  "name": "Company Name",
  "url": "https://company-website.com",
  "address": "123 Business St, City, State, ZIP"
}

Best Practices

• Query Design: Use specific, targeted search terms
• Result Processing: Always validate and enrich search results
• Rate Limiting: Respect API limits and implement delays
• Data Quality: Clean and verify company information
• Privacy: Handle company data responsibly and legally