← All modules
// PREDICTIVE MAINTENANCE

Your customer base, scheduled.

Most service businesses get sixty percent of revenue from one-time customers. Corex reads each customer's equipment + service history, forecasts the next visit, and sends a branded calendar invite when it's due — every customer, every cycle, automatically.

$149/mo · or $119/mo billed annually · live in 48 hours

// // THE PROBLEM

Tuesday morning. Half the day, empty.

Your CRM has three hundred customers you've serviced in the last two years. Your schedule has four jobs on it. The repeat is supposed to be where the money is — but the customer forgets, the owner forgets, and the next call goes to whoever they remember six months later.

60%
of revenue from one-time customers
1 in 5
past customers return on their own, without a reminder
6 months
memory window before they call someone else
// // HOW IT WORKS

Forecast. Branded reminder. One-tap booking.

1
AI reads each customer's equipment + service history and builds a prioritized maintenance forecast (critical / high / medium tasks)
2
Each task line item has a cost estimate pulled from your price book — total visit value forecast before you ever pick up a phone
3
When the next visit is due, a branded reminder lands in their inbox with a calendar invite — signed off by your team, not by Corex
4
Customer taps "Book it" — slot lands on your dispatch board automatically, with the forecast notes attached so the tech knows what to bring
5
Same loop runs every cycle, on every customer in your CRM — forever
// // WHAT YOU GET

On next week's schedule. And the next. And the next.

Yesterday, fourteen reminders fired. Six customers booked themselves before bed — twenty-four hundred dollars on next week's schedule, with zero effort from you. Run the loop on three hundred customers and a slow-week pipeline becomes a recurring-revenue machine.

$0
manual effort — the loop runs every week, every customer
~40%
of past customers typically rebook when reminded at the right time
48 hr
live in your business
// // WHO IT'S FOR

If this sounds like you — it's a fit.

HVAC shop with two slow seasons a year
Spring and fall slumps are the killers — when the phone stops ringing, the trucks sit. pm books the spring tune-ups in January and the fall maintenance in August, on autopilot.
Owner with a giant customer database and no marketing program
Three hundred names in QuickBooks, ten in active rotation. The other 290 are dormant revenue. pm wakes them up one cycle at a time.
Service biz competing with a chain on new-customer acquisition
Can't outspend the chain on Google ads — can outserve them on retention. pm makes the customer feel remembered, which is how chains lose deals to the local shop.
// // QUESTIONS

Quick answers.

What does "AI maintenance forecast" actually mean — is it making stuff up?
No. The AI reads three things: the equipment on file (make, model, age), the service history (what was done, when), and your trade's standard maintenance schedule. From those it builds a prioritized task list with cost estimates — same way a senior tech would walk a customer's history and call out what's coming due. It doesn't invent equipment or invent tasks; it organizes what your own data already tells it.
How does it know when to fire a reminder?
Each forecast has a target date based on the equipment's recommended service interval (or your trade's defaults). When the date is approaching, the reminder fires — by default 14 days before due. You can stretch or shrink the lead time per customer or per equipment type. The customer can also re-schedule themselves; you don't get pulled in unless something needs your attention.
Will customers feel spammed?
No, because they're reminders for service they actually need — not promos. The cadence is one email at the right time, not weekly newsletters. Most customers thank the shop for the reminder; the unsubscribe rate is below 2% in practice. You can also pre-set per-customer rules ("only contact about heating system," "do not contact in summer," etc.).
What if my CRM is messy or my equipment data is incomplete?
We help clean it up during setup. Most shops have enough — customer name + address + equipment type + last service date is the minimum, and pm works fine on that. The more you have (make/model/age, full service history), the sharper the forecast. We build the import pipeline during the 48-hour onboarding.

See it run on your business.

15-minute walkthrough. We'll plug in your numbers, show you what it looks like live, and tell you whether it makes sense.

Book a walkthrough →Start free trial

// 30-DAY FREE TRIAL · NO CONTRACT · 48-HOUR SETUP

// // EXPLORE MORE

Other modules

AI VOICE RECEPTIONIST
AI Voice Receptionist
PHOTO ESTIMATE & PROPOSAL GENERATOR
Photo Estimate & Proposal Generator
24/7 LEAD CLOSER
24/7 Lead Closer — Auto-Reply to Every Missed Call + Lead
AI INVOICE CHASER
AI Invoice Chaser — Friendly to Firm to Paid
See all 18 modules →