AGAI_5  /  Project 5 of 50  /  Live

REAL
ESTATE
AI DOMINATION

Autonomous lead generation system that replaces $6,000–$10,000/month in manual acquisition labor. Zero prospecting. Automated outreach on 500+ leads per day.

View Live API GitHub Repo
Scroll to explore
reds_agent — pipeline.log
$ python main.py --city houston --limit 20 ──────────────────────────────────────── ✓ Scraper initialized — Chromium headless ✓ Navigating houston.craigslist.org/rea ✓ 20 listings found. Extracting... ──────────────────────────────────────── → Claude AI scoring lead [1/20]   distress_score: 72 | signals: ["below_appraisal", "motivated_seller"] → Claude AI scoring lead [2/20]   distress_score: 18 | DISQUALIFIED → Claude AI scoring lead [3/20]   distress_score: 55 | signals: ["estate_sale", "as_is"] ──────────────────────────────────────── ✓ 7 qualified leads pushed to Supabase ✓ 7 contacts created in GHL CRM ✓ Workflow triggered: SMS + Email sequence ──────────────────────────────────────── ✓ Pipeline complete. 7 leads in outreach. $
Portfolio
AGAI_1 / WhatsApp Automation AGAI_2 / B2B Lead Generation AGAI_3 / AI Support Agent AGAI_4 / RAG Knowledge Base AGAI_5 / REDS — Real Estate AI ◆ AGAI_6 / ARA — Research Agent AGAI_1 / WhatsApp Automation AGAI_2 / B2B Lead Generation AGAI_3 / AI Support Agent AGAI_4 / RAG Knowledge Base AGAI_5 / REDS — Real Estate AI ◆ AGAI_6 / ARA — Research Agent
Monthly Labor Cost
$10,000 / mo
$50
↓ 99% reduction
Cost Per Lead
$12.00 human
$0.08
↓ 99% cheaper
Daily Lead Capacity
40–60 manual
500+
↑ 800% more volume
CRM Update Speed
15–30 min manual
<5s
↓ Sub-5-second push
The Problem

$10K/Month
in Human
Labor

Real estate acquisition teams rely on 2–3 agents spending 8 hours daily browsing listing sites, manually copying contacts into CRMs, and sending individual follow-up texts. The process is slow, inconsistent, and expensive.

REDS eliminates every manual step. The entire pipeline — from scraping to outreach — runs autonomously, 24 hours a day.

Cost Breakdown

Before
REDS

Acquisition Agent × 2 $6,000 / mo
CRM Data Entry $1,500 / mo
Follow-up Coordinator $2,000 / mo
Missed Leads (slow response) $∞ opportunity
Total Monthly Overhead $9,500+
After REDS ~$50 / mo
System Architecture

Closed-Loop
Pipeline

01
Playwright Browser Agent
Autonomous Chromium agent navigates Craigslist FSBO listings, extracting title, price, description, location, and contact details from live HTML.
Playwright Chromium Python 3.12
02
Claude AI Extraction Pipeline
Claude Sonnet analyzes raw listing text, extracts structured contact and property data, assigns a distress score from 0–100, and generates motivation summaries.
Claude Sonnet Anthropic API Pydantic
03
Supabase Storage + Deduplication
Qualified leads are stored in PostgreSQL with URL-based deduplication. Prevents duplicate outreach across scrape cycles.
Supabase PostgreSQL RLS Policies
04
GoHighLevel CRM Integration
Pushes contacts to GHL via API v2 with distress tags, AI analysis notes, and custom fields. Contact creation fires the outreach workflow automatically.
GHL API v2 Contact Tags httpx async
05
Automated Outreach Sequences
GHL workflow fires on REDS_Lead tag: SMS opener, 1-day wait, email follow-up, 2-day wait, SMS second touch. Zero manual intervention.
GHL Workflows SMS Email
06
Reply Monitor + FastAPI Dashboard
Polls GHL conversations for inbound replies, updates lead status in Supabase. FastAPI exposes /leads, /metrics, and /health endpoints for full pipeline visibility.
FastAPI Uvicorn structlog Railway
Technology

Built with
Precision

Scraping
Playwright
Headless Chromium automation with verified live selectors. Polite delay between requests. City-level targeting.
AI / LLM
Claude AI
Claude Sonnet for lead extraction, distress scoring, and motivation analysis. JSON-structured outputs with Pydantic validation.
CRM
GoHighLevel
API v2 contact creation with custom fields, tags, and notes. Contact tag workflow trigger for outreach automation.
Database
Supabase
PostgreSQL with RLS policies, unique URL index for deduplication, and status-based lead filtering.
API Layer
FastAPI
Async REST API with Swagger UI. Endpoints for leads, metrics, and health. Structured JSON logging via structlog.
Deployment
Railway
Docker container deployment via Microsoft Playwright base image. Auto-deploy on GitHub push. Custom domain at reds.datawebify.com.
AI Scoring System

Distress
Detection

"must sell" / "motivated seller"
+25
Foreclosure / pre-foreclosure
+30
Divorce / estate sale / probate
+25
Price reduced / below appraisal
+20
As-is / needs work / fixer upper
+20
Relocating / job transfer
+15
Vacant / empty property
+15
Days on market over 60
+15
Cash only / no realtor / FSBO
+10
Asking below $150,000
+10
Sample Lead Score
72
✓ QUALIFIED — Entering GHL pipeline

Score ≥ 30 → Qualified, pushed to GHL
Score < 30 → Filtered, stored as disqualified

Signals detected:
"below appraisal" +20
"motivated seller" +25
"cash discount" +10
FSBO listing +10
Asking < $150K +10
Live API

REST
Endpoints

Full pipeline visibility via FastAPI. Live at reds.datawebify.com with interactive Swagger documentation.

GET /health System health check
GET /leads All leads, optional ?status= filter
GET /leads/{id} Single lead by UUID
GET /metrics Pipeline metrics and business impact
GET /docs Interactive Swagger UI

Deploy This
For Your
Business

$15,000 — $40,000 per engagement  /  datawebify.com

Get in Touch Explore API
Chat on WhatsApp