AI-Powered Contact Intelligence & Enrichment with OpenAI/Anthropic and Supabase

AI Contact Enrichment

📋 Template Description

Overview Automatically enhance and enrich contact data using AI to fill in missing information, generate insights, and create detailed buyer personas. Supports multiple AI providers (OpenAI, Anthropic, etc.) with automatic logging to Supabase.

Description This workflow transforms incomplete contact records into rich, actionable profiles. By leveraging AI, it can infer job roles, company information, likely pain points, communication preferences, and buying motivations from minimal input data. Perfect for sales and marketing teams looking to improve data quality and personalize outreach.

Key Benefits: Smart Data Completion**: Fill in missing contact fields using AI inference Buyer Persona Generation**: Create detailed profiles from basic information Universal AI Support**: Works with OpenAI, Anthropic Claude, or custom providers CRM Enhancement**: Automatically enrich contacts as they enter your system Lead Qualification**: Assess lead quality and fit based on enriched data Personalization Engine**: Generate insights for tailored outreach Data Quality**: Maintain clean, complete contact records

Use Cases: Sales prospecting and lead enrichment Marketing persona development CRM data cleansing and completion Account-based marketing (ABM) research Lead scoring and qualification Personalized email campaign preparation Contact segmentation and targeting

⚙️ Setup Instructions

Prerequisites n8n instance (cloud or self-hosted) AI Provider account (OpenAI, Anthropic, or custom) Supabase account with database access

Step 1: Configure Environment Variables Add these to your n8n environment settings:

AI_PROVIDER=openai # or 'anthropic', 'custom' AI_API_KEY=your_api_key_here AI_MODEL=gpt-3.5-turbo # or 'gpt-4', 'claude-3-sonnet-20240229' AI_ENDPOINT= # Only for custom providers

Recommended Models: Cost-effective**: gpt-3.5-turbo (fast, affordable, good for basic enrichment) High-quality**: gpt-4 or claude-3-sonnet-20240229 (better inference, deeper insights) Premium**: claude-3-opus-20240229 (best for complex persona generation)

How to set environment variables: n8n Cloud**: Go to Settings → Environment Variables Self-hosted**: Add to your .env file or docker-compose configuration

Step 2: Set Up Supabase Database Create the logging table in your Supabase database:

CREATE TABLE workflow_logs ( id BIGSERIAL PRIMARY KEY, workflow_name TEXT NOT NULL, data JSONB NOT NULL, ai_response JSONB NOT NULL, created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() );

CREATE INDEX idx_workflow_logs_created_at ON workflow_logs(created_at); CREATE INDEX idx_workflow_logs_workflow_name ON workflow_logs(workflow_name);

-- Optional: Create a view for enriched contacts CREATE VIEW enriched_contacts AS SELECT id, data->>'email' as email, data->>'name' as name, data->>'company' as company, ai_response as enrichment_data, created_at FROM workflow_logs WHERE workflow_name = 'AI Contact Enrichment' ORDER BY created_at DESC;

To run this SQL: Open your Supabase project dashboard Go to the SQL Editor Paste the SQL above and click "Run"

Step 3: Configure Supabase Credentials in n8n Go to Settings → Credentials Click Add Credential → Supabase API Enter your Supabase URL and API key (found in Project Settings → API) Name it Supabase API Click Save

Step 4: Activate the Webhook Import this workflow into n8n Click the Activate toggle in the top-right corner Click on the "Webhook Trigger" node Copy the Production URL (this is your webhook endpoint) Save this URL for integration with your applications

Step 5: Test the Workflow Send a test POST request to the webhook:

curl -X POST https://your-n8n-instance.com/webhook/contact-enrichment
-H "Content-Type: application/json"
-d '{ "email": "john.doe@acmecorp.com", "name": "John Doe", "company": "Acme Corporation", "linkedin_url": "https://linkedin.com/in/johndoe" }'

Successful Response: { "success": true, "workflow": "AI Contact Enrichment", "timestamp": "2025-01-14T12:00:00.000Z" }

📥 Expected Payload Format

The webhook accepts JSON with basic contact information:

Minimal Input { "email": "string (required or name required)", "name": "string (required or email required)" }

Recommended Input { "email": "string", "name": "string", "company": "string", "job_title": "string", "linkedin_url": "string", "phone": "string", "location": "string", "website": "string" }

Complete Input Example { "email": "sarah.chen@techstartup.io", "name": "Sarah Chen", "company": "TechStartup Inc.", "job_title": "VP of Marketing", "linkedin_url": "https://linkedin.com/in/sarahchen", "phone": "+1-555-0123", "location": "San Francisco, CA", "website": "https://techstartup.io", "industry": "B2B SaaS", "company_size": "50-200 employees", "notes": "Met at SaaS conference 2024" }

Field Guidelines: At minimum, provide either email or name More input fields = better AI enrichment quality Include linkedin_url for best results company helps with firmographic enrichment Any additional context improves accuracy

🔄 Workflow Flow

Webhook Trigger: Receives basic contact information from your application, form, or CRM Process Data: Adds unique ID and timestamp to the incoming data Prepare AI Request: Configures AI provider settings from environment variables Call AI API: Sends contact data to AI with enrichment prompt Save to Supabase: Archives original data and enrichment results Format Response: Returns success confirmation

🎯 Customization Tips

Enhance AI Prompts for Better Enrichment

Modify the "Prepare AI Request" node to customize enrichment:

// Enhanced prompt for contact enrichment const systemPrompt = `You are an expert sales intelligence analyst. Analyze the provided contact information and generate a comprehensive enrichment including:

INFERRED DETAILS: Fill in missing information based on available data Full job title and seniority level Department and reporting structure Years of experience (estimated) Professional background

COMPANY INSIGHTS: If company name provided Industry and sub-industry Company size and revenue (estimated) Key products/services Recent news or developments

BUYER PERSONA: Create a detailed profile Primary responsibilities Likely pain points and challenges Key priorities and goals Decision-making authority Budget influence level

ENGAGEMENT STRATEGY: Provide outreach recommendations Best communication channels Optimal outreach timing Key talking points Personalization suggestions Content interests

LEAD SCORE: Rate 1-10 based on: Fit for product/service (specify your ICP) Seniority and decision power Company size and maturity Engagement potential

Return as structured JSON with clear sections.`;

const userMessage = Contact Information:\n${JSON.stringify($json.data, null, 2)};

const aiConfig = { provider: $env.AI_PROVIDER || 'openai', apiKey: $env.AI_API_KEY, model: $env.AI_MODEL || 'gpt-3.5-turbo', endpoint: $env.AI_ENDPOINT, messages: [ { role: 'system', content: systemPrompt }, { role: 'user', content: userMessage } ] };

return { json: { aiConfig, data: $json } };

Add External Data Sources

Enhance enrichment with third-party APIs:

After "Process Data" node, add:

Clearbit/Hunter.io Node: Get verified company data LinkedIn API: Pull professional information Company Database: Query internal customer data Web Scraping: Extract data from company websites

Then merge all data before AI enrichment for best results

Connect to Your CRM

Auto-update contacts after enrichment:

Salesforce Integration: // Add after "Call AI API" node // Update Salesforce contact with enriched data const enrichedData = JSON.parse($json.ai_response); return { json: { contactId: $json.data.salesforce_id, updates: { Description: enrichedData.buyer_persona, Custom_Score__c: enrichedData.lead_score, Pain_Points__c: enrichedData.pain_points } } };

HubSpot Integration: Add HubSpot node to update contact properties Map enriched fields to custom HubSpot properties

Pipedrive Integration: Use Pipedrive node to update person records Add custom fields for AI insights

Implement Lead Scoring

Add scoring logic after enrichment:

// Calculate lead score based on enrichment const enrichment = JSON.parse($json.ai_response);

let score = 0;

// Job title scoring if (enrichment.seniority === 'C-Level') score += 30; else if (enrichment.seniority === 'VP/Director') score += 20; else if (enrichment.seniority === 'Manager') score += 10;

// Company size scoring if (enrichment.company_size === 'Enterprise') score += 25; else if (enrichment.company_size === 'Mid-Market') score += 15;

// Decision authority scoring if (enrichment.decision_authority === 'High') score += 25; else if (enrichment.decision_authority === 'Medium') score += 15;

// Budget influence if (enrichment.budget_influence === 'Direct') score += 20;

return { json: { ...enrichment, lead_score: score } };

Add Compliance Checks

Insert before AI processing:

// Check for opt-out or compliance flags const email = $json.email.toLowerCase();

// Check against suppression list const suppressedDomains = ['competitor.com', 'spam.com']; const domain = email.split('@')[1];

if (suppressedDomains.includes(domain)) { throw new Error('Contact on suppression list'); }

// Verify email format const emailRegex = /^+@+.+$/; if (!emailRegex.test(email)) { throw new Error('Invalid email format'); }

return { json: $json };

Batch Enrichment

Process multiple contacts:

Add Spreadsheet File trigger instead of webhook Add Split In Batches node (process 10-20 at a time) Run enrichment for each contact Combine results and export to CSV

🛠️ Troubleshooting

Common Issues

Issue: "Enrichment is too generic" Solution**: Provide more input data (company, job title, LinkedIn) Use GPT-4 or Claude models for better inference Enhance the system prompt with specific instructions

Issue: "AI_API_KEY is undefined" Solution**: Ensure environment variables are set correctly Verify variable names match exactly (case-sensitive)

Issue: "Enrichment contradicts actual data" Solution**: AI makes inferences - always validate critical information Add validation step to check enriched data against known facts Use external APIs for verification

Issue: "Too slow for real-time use" Solution**: Implement queue system for async processing Use faster models (gpt-3.5-turbo) for speed Process in batches during off-peak hours

Issue: "Supabase credentials not found" Solution**: Check credential name matches exactly: "Supabase API" Verify Supabase URL and API key are correct

Debugging Tips Test with known contacts first to validate accuracy Compare AI enrichment against actual data Check execution logs for API errors Start with minimal prompt, then enhance gradually Use "Execute Node" to test individual steps

📊 Analyzing Enriched Data

Query and analyze your enriched contacts:

-- Get all enriched contacts SELECT * FROM enriched_contacts ORDER BY created_at DESC;

-- Find high-value leads (assuming scoring implemented) SELECT email, name, company, ai_response->>'lead_score' as score FROM enriched_contacts WHERE (ai_response->>'lead_score')::int > 70 ORDER BY (ai_response->>'lead_score')::int DESC;

-- Analyze enrichment by company SELECT data->>'company' as company, COUNT(*) as contact_count, AVG((ai_response->>'lead_score')::int) as avg_score FROM workflow_logs WHERE workflow_name = 'AI Contact Enrichment' AND ai_response->>'lead_score' IS NOT NULL GROUP BY data->>'company' ORDER BY contact_count DESC;

-- Find contacts needing follow-up SELECT email, name, ai_response->>'engagement_strategy' as strategy, created_at FROM enriched_contacts WHERE created_at > NOW() - INTERVAL '7 days' ORDER BY created_at DESC;

Export Enriched Data

-- Export to CSV COPY ( SELECT data->>'email' as email, data->>'name' as name, data->>'company' as company, ai_response->>'job_title' as enriched_title, ai_response->>'seniority' as seniority, ai_response->>'lead_score' as score FROM workflow_logs WHERE workflow_name = 'AI Contact Enrichment' ) TO '/tmp/enriched_contacts.csv' WITH CSV HEADER;

📈 Integration Ideas

Form Integration Automatically enrich new leads from forms: Typeform**: Trigger on form submission Google Forms**: Use Google Sheets trigger Calendly**: Enrich after meeting booking Webflow Forms**: Webhook trigger from form

CRM Integration Real-time enrichment as contacts enter CRM: Salesforce**: Trigger on new lead/contact creation HubSpot**: Enrich on form submission or import Pipedrive**: Auto-enrich new persons Close**: Webhook on lead creation

Email Tools Enhance cold outreach campaigns: Instantly.ai**: Enrich before campaign launch Lemlist**: Generate personalization variables Apollo.io**: Supplement with AI insights Mailshake**: Enrich prospect lists

Marketing Automation Power ABM and segmentation: Marketo**: Enrich leads for scoring Pardot**: Enhance prospect profiles ActiveCampaign**: Personalization data Klaviyo**: E-commerce customer insights

Slack Integration Team notifications and collaboration: Send enrichment summaries to sales channel Notify reps of high-value leads Share persona insights with marketing Alert on key account contacts

🔒 Security & Compliance Best Practices

Data Protection Encrypt Sensitive Data: Use environment variables for all credentials Access Control: Limit webhook access with authentication Data Retention: Set automatic deletion policies in Supabase Audit Logging: Track all enrichment activities

Privacy Compliance GDPR Compliance: Get consent before enriching personal data Allow contacts to request data deletion Document legal basis for processing CCPA Compliance: Honor do-not-sell requests Data Minimization: Only enrich necessary fields Right to Access: Allow contacts to view enriched data

AI Ethics Bias Awareness: Review AI inferences for bias Accuracy Validation: Verify critical information Transparency: Disclose use of AI enrichment Human Oversight: Review before critical decisions

💡 Best Practices

Input Data Quality Always include email or full name** as anchor point Add LinkedIn URLs** for 50% better accuracy Provide company name** for firmographic insights Include any known details** - more data = better results

Prompt Engineering Be specific** about your ideal customer profile (ICP) Request structured output** (JSON format) Define scoring criteria** that match your business Ask for actionable insights** not just descriptions

Post-Enrichment Workflow Always validate** critical information before use Review AI inferences** for accuracy and bias Update CRM promptly** to maintain data freshness Track enrichment ROI** (conversion rates, time saved)

Performance Optimization Batch process** during off-peak hours Use appropriate models** (gpt-3.5 for speed, gpt-4 for quality) Cache common enrichments** to reduce API costs Set rate limits** to avoid API throttling

🏷️ Tags sales-automation, lead-enrichment, ai-automation, crm-integration, data-enrichment, contact-intelligence, buyer-personas, lead-scoring, webhook, supabase, openai, anthropic, b2b-sales

📝 License This workflow template is provided as-is for use with n8n.

🤝 Support For questions or issues: n8n Community Forum: https://community.n8n.io n8n Documentation: https://docs.n8n.io

🌟 Example Output

Input: { "email": "mike.johnson@cloudtech.com", "name": "Mike Johnson", "company": "CloudTech Solutions", "job_title": "Director of IT" }

AI-Generated Enrichment: { "full_title": "Director of Information Technology", "seniority": "Director", "department": "Technology/IT", "experience_years": "10-15", "company_insights": { "industry": "Cloud Computing", "size": "Mid-Market (100-500)", "revenue_estimate": "$10M-$50M" }, "buyer_persona": { "responsibilities": ["Infrastructure management", "Vendor selection", "Security oversight"], "pain_points": ["Legacy system migration", "Cost optimization", "Security compliance"], "priorities": ["Scalability", "Cost reduction", "Team efficiency"] }, "engagement_strategy": { "best_channels": ["Email", "LinkedIn"], "timing": "Tuesday-Thursday, 9-11 AM", "talking_points": ["ROI and cost savings", "Security features", "Ease of implementation"], "personalization": "Reference cloud migration challenges" }, "lead_score": 75 }

🔄 Version History v1.0.0** (2025-01-14): Initial release with universal AI provider support