AI-Powered Lead Generation System for Marketing Agency
Azen AI
Industry: Local Businesses, Agencies, B2B Services
Solution: Automated Lead Sourcing, Research, AI Personalization, Email Outreach
Tech: n8n, OpenAI, Google Sheets, Apify, Telegram
Overview
A boutique marketing agency wanted a way to scale outbound lead generation without hiring SDRs or spending hours manually researching prospects.
Their biggest challenges were finding high-quality leads consistently, collecting data from LinkedIn and websites, personalizing every email to avoid sounding generic, sending timely outreach without manual effort, and keeping everything neatly updated in one database.
They wanted the entire system to run on autopilot.
So we built them a three-stage AI-powered lead generation engine inside n8n that finds leads through automated scraping and qualification, personalizes emails using AI based on LinkedIn and website insights, and sends automated warm outreach with tracking.
Everything runs from one central workflow—with the option to trigger steps from Telegram for real-time control.
Stage One: Automated Lead Discovery
The workflow starts with a scheduled trigger that runs daily.
How it works:
ICP Search (Ideal Customer Profile)
The system uses predefined parameters including industry, location, business type, employee size, and keywords. These parameters are stored in a Config node, allowing the agency to change its target niche anytime.
Apify LinkedIn Scraper
We built a clean pipeline that sends ICP parameters to Apify, runs a scraper session, waits automatically for job completion, and fetches final results.
Lead Conditioning and Deduplication
The system checks whether each lead is new, has a valid website, and represents an active company.
Write Leads to Database
Leads are written into Google Sheets (or Airtable or Notion depending on client preference).
Time Saved: What used to take four to five hours per day of manual searching is now done in under 60 seconds.
Stage Two: AI-Powered Email Personalization
This is where the magic happens. Once leads are added, the workflow moves to deep research and AI personalization.
The research pipeline includes:
LinkedIn scraping that captures headline, bio, posts, experience, and activity. Website scraping covers homepage, services, and about page. The system performs keyword extraction, tone modeling, and offer relevance scoring.
Then the Personalization Agent—a custom OpenAI agent—combines all gathered information including prospect background, company description, what they sell, their recent activity, pain points found on their site, and how your offer can fit them.
The workflow generates a fully personalized email that feels handcrafted.
Example output generated by the AI agent:
The email includes a direct reference to their latest project, a compliment based on their recent website updates, a pain-point analysis tied to what they offer, a custom offer suggestion from the configured offers section, and a friendly human-sounding opener.
Result: Emails feel uniquely tailored, not AI-generated or template-based. This stage increased clients’ positive reply rates by three times.
Stage Three: Automated Outreach
Once the email is ready, the workflow sends it through an automated email pipeline.
How sending works:
Scheduled Daily Outreach
Every morning the workflow selects qualified leads.
Validation and Rate Controls
The system performs email verification, throttled sending to protect domain reputation, and duplicate prevention.
Send Email with Resend Node
The email includes the prospect’s name, custom subject line, personalized email body, and optional CTA button.
Write Back to CRM
The system logs email sent status, timestamp, message content, and lead status.
Warm follow-up nodes ready to plug in
End-to-end automated. Even follow-ups can be added without writing a single line of code.
Key Workflow Highlights
Fully automated lead generation: Daily scraping, qualification, and enrichment with zero manual effort.
Human-level personalized emails: AI agents generate emails that sound like a real SDR wrote them.
Modular structure: Change niches, offers, or research sources instantly.
Telegram control: Start, stop, or test the workflow directly from Telegram.
Scalable and predictable: Run daily, weekly, or on-demand without additional cost.
Results
Time spent finding leads: Previously four to five hours per day, now zero minutes.
Email personalization time: Previously three to seven minutes per email, now under 10 seconds.
Warm leads added per month: Previously approximately 150, now 900 plus.
Positive reply rate: Previously 1.8%, now 5.9%.
Labor cost: Previously high, now zero additional cost.
What the Client Said
“This feels like having a full-time SDR team working in the background. We went from struggling to find good leads to having more appointments than we can handle.”
This automation shows what’s now possible when scraping, AI research, and email sending are stitched together inside a single n8n workflow.
Instead of juggling tools, Chrome tabs, freelancers, and manual research, you get one automated system that finds ideal leads, researches them like a human, writes personalized messaging, sends emails automatically, updates your CRM, and runs every day without failing.
This is the future of outbound for small teams and local businesses.
Book a free automation audit and discover how we can build a similar system for your business.
Like this project
Posted Apr 25, 2026
Built an AI-powered lead generation engine to automate lead sourcing and outreach for a marketing agency.
Likes
0
Views
1
