AI-Powered LinkedIn Lead Generation Platform — Autonomous Outreach System Built with n8n
I built a fully autonomous LinkedIn Lead Generation Platform that scrapes
targeted prospects, scores them with AI, sends hyper-personalized connection
requests and DMs, detects replies, runs multi-step follow-up sequences, and
delivers daily analytics reports — all without any manual intervention.
This end-to-end system is powered by n8n, OpenAI GPT-4, Unipile, Apify,
Google Sheets, Telegram, and Gmail. It operates 24/7, respects LinkedIn’s
rate limits with human-like delays, supports multiple LinkedIn accounts,
and automatically updates a CRM in real time.
The result: a scalable, intelligent outreach engine that replaces hours of
manual LinkedIn work with a fully automated pipeline.
The platform is built as 6 interconnected n8n workflows, each handling a specific stage of the LinkedIn outreach funnel — from lead discovery to
conversion tracking. Every workflow communicates through Google Sheets as a shared CRM layer, with Telegram and Gmail providing real-time reporting.
6 Automated Workflows
AI-Powered Personalization
Multi-Account Support
Real-Time CRM Updates
Zero Manual Effort
WF1 — LinkedIn Lead Scraping
Trigger: Scheduled (runs automatically at set intervals)
Tools: Apify, Google Sheets, Telegram
The first workflow is responsible for discovering and collecting targeted
LinkedIn profiles. A Schedule Trigger fires Apify’s LinkedIn scraper, which
extracts prospect data based on predefined search criteria (industry,
job title, location, etc.).
The scraped results are fetched, cleaned, and formatted into a structured
dataset. Before saving, the system checks Google Sheets for duplicates —
only genuinely new leads pass through. Each new lead is saved to the master Google Sheet CRM, a Telegram alert is triggered instantly, and the analytics tracker is updated.
Key Nodes: Schedule Trigger → Apify Scrape → Fetch Results → Clean & Format → Duplicate Check → Is New Lead? → Save to Google Sheets → Telegram Alert → Update Analytics
WF2 — AI Personalized DM Generator
Trigger: Scheduled
Tools: OpenAI GPT-4, Google Sheets
Once leads are collected, this workflow prepares hyper-personalized outreach messages before any message is sent. Leads ready for outreach are fetched from Google Sheets and split into manageable batches for processing.
Each lead goes through an AI Lead Scoring step, which evaluates profile
relevance. High-scoring leads are passed to OpenAI GPT-4, which generates a unique, context-aware connection note or DM tailored to the prospect’s profile. The generated messages are parsed, validated for quality and length, then saved back to Google Sheets — ready for the sending workflow. A rate
limit delay prevents API overuse.
Key Nodes: Schedule Trigger → Get Leads → Split Into Batches → AI Lead Scoring → OpenAI Generate DMs → Parse & Validate → Save to Google Sheets → Rate Limit Delay
WF3 — Unipile LinkedIn Messaging Automation
Trigger: Hourly
Tools: Unipile API, Google Sheets
This is the core sending engine. It runs every hour and intelligently
manages the actual delivery of connection requests and direct messages
across multiple LinkedIn accounts.
The workflow first checks business hours — outreach only runs during
defined working hours to mimic human behavior. It then fetches active
LinkedIn accounts and pending leads, applies daily sending limits per
account, and rotates between accounts to avoid LinkedIn restrictions.
Each lead is processed one at a time. If the prospect is already connected,
a DM is sent via Unipile and the CRM is updated as “DM Sent.” If not yet
connected, a personalized connection request is sent and logged as
“Connection Sent.” A randomly calculated human-like delay is added between each action to avoid bot detection.
Key Nodes: Hourly Trigger → Business Hours Check → Active Accounts → Pending Leads → Daily Limits & Account Rotation → Process One Lead at a Time → Already Connected? → Send DM / Send Connection Request → Update CRM → Random Human-Like Delay
WF4 — Reply Detection & CRM Update
Trigger: Every 30 Minutes
Tools: Unipile API, OpenAI GPT-4, Google Sheets, Telegram
Once messages are sent, this workflow continuously monitors LinkedIn
conversations for replies. Every 30 minutes, it fetches recent conversations from Unipile for all active accounts, extracts new replies, and matches them against the messaged leads in Google Sheets. Each reply is sent to OpenAI for intent analysis — the AI classifies the reply as Interested, Not Interested, or Neutral. The CRM is updated with the reply status and sentiment. If a lead is identified as a Hot Lead (showing buying intent), an instant Telegram alert is fired so the opportunity can be followed up on immediately. Reply counts in the analytics sheet are also updated.
Key Nodes: Poll Every 30 Min → Active Accounts → Fetch Recent Conversations → Extract New Replies → Match Reply to Lead → OpenAI Analyze Reply → Parse Intent → Update CRM with Reply Status → Is Hot Lead? → Telegram Alert →Update Analytics
WF5 — Follow-up Automation
Trigger: Daily
Tools: Unipile API, Google Sheets
Most LinkedIn outreach fails because follow-ups don’t happen consistently. This workflow solves that completely. Every day, it fetches all leads who have not replied and classifies them by follow-up stage (Follow-up 1 or Follow-up 2). Each lead is processed individually. Based on which follow-up stage they’re in, a context-appropriate message is prepared (FU1 is softer; FU2 creates urgency). The correct LinkedIn account is retrieved, the follow-up is sent via Unipile, and the CRM is updated to reflect which follow-up was sent. A random delay before the next lead ensures natural, human-like behavior.
Key Nodes: Daily Trigger → Get Leads With No Reply → Classify Follow-up Type → One Lead at a Time → Which Follow-up? → Prepare FU1 / FU2 Message → Get LinkedIn Account → Unipile Send Follow-up → Update CRM → Random Delay → Wait
WF6 — Analytics Dashboard & Daily Reporting
Trigger: Daily
Tools: Google Sheets, Telegram, Gmail
Every day, this workflow compiles the complete performance data of the
entire LinkedIn outreach system. It reads all lead records from Google
Sheets, calculates key metrics (leads scraped, messages sent, connection
rate, reply rate, hot leads identified), and writes a structured analytics
summary to a dedicated analytics sheet. A formatted daily report is sent via Telegram for instant mobile access, and a detailed email report is dispatched via Gmail for documentation and client reporting.
Key Nodes: Daily Analytics Trigger → Read All Leads → Calculate Analytics → Write to Analytics Sheet → Telegram Daily Report → Gmail Daily Report Email
Tech Stack Used
n8n — Workflow orchestration & automation engine
OpenAI GPT-4 — AI lead scoring & personalized message generation
Unipile API — LinkedIn DM & connection request delivery
Apify — LinkedIn profile scraping
Google Sheets — Shared CRM, data storage & analytics layer
Telegram Bot API — Real-time alerts & daily reports
Gmail API — Daily email reporting
Key Features
✦ Fully autonomous — runs 24/7 with zero manual input
✦ AI-generated, hyper-personalized messages for every prospect
✦ Smart duplicate prevention — no lead is contacted twice
✦ Multi-account LinkedIn support with daily limit rotation
✦ Human-like random delays to avoid LinkedIn bot detection
✦ Business-hours-only sending for realistic outreach patterns
✦ Automated 2-stage follow-up sequences
✦ Real-time reply detection with AI sentiment analysis
✦ Hot lead instant Telegram alerts
✦ Daily analytics via Telegram + Gmail
✦ Google Sheets as a lightweight, shareable CRM
Results & Impact
This system transforms LinkedIn from a manual, time-consuming channel into a fully automated lead generation machine. Key outcomes include:
— Eliminates 3–5 hours of daily manual outreach work
— Ensures 100% follow-up consistency with no leads falling through the cracks
— AI personalization increases reply rates compared to generic templates
— Hot lead alerts enable immediate response to high-intent prospects
— Multi-account rotation maximizes daily outreach volume safely
— Daily reporting gives complete visibility into pipeline performance
Use Cases
This platform is ideal for:
→ Freelancers & consultants scaling their client acquisition
→ B2B SaaS companies running outbound sales campaigns
→ Recruiters sourcing candidates at scale
→ Digital agencies managing LinkedIn outreach for multiple clients
→ Entrepreneurs building their network and pipeline autonomously
Want This System for Your Business?
I build custom AI-powered LinkedIn automation systems tailored to your industry, target audience, and outreach goals.