AI-powered lead qualification and CRM automation system for a real estate investment project.

Context

Business: Nova Center — a real estate investment project selling to private investors.

Scale (approx.): ~80 inquiries per week from a landing page, handled by one administrator.

Before: leads came from a Wix form, were copied by hand into a spreadsheet and forwarded to managers in messengers. There was no lead qualification and no CRM workflow.

Goal: process every lead instantly, show managers which leads are hot, and keep all data in one place.

ChatGPT Image 28 квіт. 2026 р., 12_43_46.png


Problem

— Manual lead processing consumed significant time

— Leads could be missed or processed too slowly

— No lead qualification system existed

— Managers spent time on low-quality leads

— Data was scattered across multiple tools


Solution

I built an AI-powered automation system using n8n, AI models and CRM integrations.

The system: — captures leads automatically — analyzes lead intent using AI — classifies leads as hot / warm / cold — sends notifications to Telegram — stores data in Google Sheets and Zoho CRM — routes contacts into segmented Klaviyo lists — scores each lead by signals such as stated budget, urgency and message intent, combining them into a qualification score


Process

  1. Discovery — mapped how a lead moved from the website form to a manager and where time was lost (manual copying, no priorities)
  2. Solution design — defined hot / warm / cold criteria (budget, urgency, intent) and where each type of lead should go
  3. Build — created the n8n workflow: form → AI qualification → CRM, Google Sheets, Telegram and Klaviyo
  4. Testing — ran test leads and compared AI decisions with a manual review of the same leads
  5. Launch — connected the live landing page form and added retry / error handling so no lead is lost
  6. Optimization — refined the qualification prompt and scoring signals based on test results