Stop chasing leads.
Let AI find contractors who need you.
APEX Full Sales Machine scrapes Google Maps and CSLB licensing daily, scores every lead by fit, researches each prospect, sends personalized outreach, calls with AI voice, and books the meeting. $0.05 per qualified lead. Proven in production.
Contractor lead gen is broken. Here's why the old ways fail.
Lead marketplaces sell the same lead to five competitors. Cold lists are stale. A sales rep costs $25/hr and manually works 20 leads per day. None of it scales.
Lead marketplaces sell shared leads
Angi, HomeAdvisor, and Thumbtack charge $15–$80 per lead and sell the same contact to multiple contractors simultaneously. You're in a race to call first, every time.
Sales reps are expensive and inconsistent
A $25/hr sales hire manually works 20 leads per day — roughly $2.50 per contact attempt before wages, benefits, and turnover. Their output depends on their mood. APEX runs the same perfect script every night regardless.
Generic lists lack context
Purchased lists have company name and phone. APEX adds license status, years in business, Google review velocity, ratings trajectory, and a GPT-generated pitch angle — before any outreach fires.
Seven layers. From raw data to booked meeting.
Every layer runs autonomously each night. No manual intervention between scrape and booking.
Scrape the market
Google Maps + CSLB scraped across your target cities daily. Dedup against existing CRM contacts. Geo-fenced to your service area.
Google Maps CSLB
Qualify by fit
License status, review velocity, business age, and ICP criteria collapse 1,000+ scraped businesses to ~100 hot leads. Ranked by probability to convert.
XGBoost SQLite
Enrich each prospect
GPT-4o-mini reads their website, recent reviews, competitor positioning, and license filings. Outputs a five-line pitch angle unique to each lead.
GPT-4o-mini Brave
Build the pitch
Custom email sequence and landing page rendered per lead. Their logo, their reviews, your offer. Hosted on a tokenized URL for tracking.
FFmpeg Nginx
Multi-touch outreach
3-touch email sequence with A/B variants. Reply detection pauses the sequence on a positive signal. CAN-SPAM compliant.
SMTP Snov.io
AI voice follow-up
Vapi.ai voice agent calls every hot lead. TCPA-compliant DNC scrub, AI identification per CA SB 1001, recording disclosure, live opt-out detection.
Vapi DNC
Land the meeting
Qualified prospects get a Cal.com link mid-call. Booking posts to CRM, fires Telegram alert to you, and stops all further outreach.
Cal.com Telegram
# 2026-04-11 03:14 PT — nightly run, electrical, san-diego
$ python apex/pipeline.py --vertical electrical --city san-diego
[01] discover scraped 247 businesses 2.4s
[02] score qualified 31 hot, 89 warm 0.6s
[03] research enriched 31 with GPT context 14.2s
[04] personalize rendered 31 video decks 62.0s
[05] pitch queued 31 sequences (3-touch A/B) 0.3s
[06] call scheduled 31 voice followups 0.1s
[07] book 4 meetings captured overnight 8h
→ 4 booked meetings | $1.74 total cost | $0.056/lead | done.
Real numbers from the live San Diego pipeline.
This is not a demo environment. These numbers come from nightly production runs against the San Diego trades market.
Questions contractors ask about AI lead gen.
How does AI lead generation work for contractors?
APEX runs a seven-layer pipeline each night. Layer 1 scrapes Google Maps and state contractor registries (CSLB for California) across your target cities, collecting business name, phone, address, owner name, license number, review count, and rating. Layer 2 scores each prospect against your ideal customer profile using license status, business age, review velocity, and growth signals — collapsing 1,000+ scraped businesses to roughly 100 hot leads. Layers 3 through 7 research each hot lead, write a personalized pitch, send multi-touch email outreach, call with a TCPA-compliant AI voice agent, and book qualified prospects directly into your calendar.
How much does AI lead generation cost for contractors?
The Full Sales Machine starts at $5,000 for a custom deployment. In production, the per-lead qualification cost is $0.05 — that is what APEX spends in API and compute costs to scrape, score, and research a lead with GPT-4o-mini. Compare that to Angi or HomeAdvisor charging $15 to $80 per lead for unqualified inbound traffic, or a $25/hr sales rep who manually works 20 leads per day at roughly $2.50 per lead before wages, benefits, and churn.
What contractor types does APEX target?
APEX is calibrated for licensed contractors across the major trade categories: general contractors (Class B), electrical (C-10), plumbing (C-36), HVAC (C-20), roofing (C-39), concrete (C-8), painting (C-33), landscaping (C-27), and specialty trades. The scoring model pulls from CSLB license classifications in California and adapts to equivalent licensing boards in other states. If you serve a niche trade, we configure the ICP scoring accordingly during onboarding.
How long until I see leads from APEX?
The pipeline runs its first nightly scrape-and-score within 24 hours of deployment. You will see a scored lead report the morning after go-live. Outreach sequences begin firing within 48 hours. First booked meetings typically land within 7 to 14 days depending on your market density and ICP definition. The Full Sales Machine deployment itself takes 14 to 30 days depending on integrations and geographic scope.
Is APEX TCPA compliant for contractor outreach?
Yes. Every outbound call scrubs against the federal Do Not Call registry before dialing. The AI voice agent identifies itself as an automated assistant in the opening sentence, per California SB 1001. Recording disclosure is delivered at the start of every call. Any prospect who says "do not call", "stop calling", or "remove me" triggers an immediate DNC flag — the lead status flips to do_not_contact and the phone number is added to our internal DNC list. Email sequences are CAN-SPAM compliant with unsubscribe footers on every message.
Request a live demo with your vertical.
We'll run a sample scrape-and-score against your target market and show you the actual lead output before you commit. No slideware. Real data from your city.