Setup checklist for an HVAC AI receptionist
A complete checklist for deploying the 365agents HVAC Dispatch agent — service area mapping, dispatch integration, emergency protocols, technician rules, and pricing configuration.
Written By Catherine Weir
Last updated 9 days ago
Deploying an AI receptionist for an HVAC or home services business typically takes about a week of elapsed time. The most important setup work is mapping your service area correctly, integrating with your field service software, and configuring emergency protocols that respect the safety-critical nature of the work.
Before the kickoff call
Field service software — which one (ServiceTitan, Housecall Pro, Jobber, FieldEdge, Service Fusion, RazorSync, WorkWave) and admin credentials
Service area definition — zip codes, cities, or a mapped polygon you currently use; any exceptions or partial-coverage areas
Hours — including after-hours rates if applicable
Trades covered — HVAC only, or HVAC plus plumbing, electrical, etc.
Technician roster — who's on what territory, who's available after hours, specialties
Diagnostic fee and service rates — standard visit, after-hours, weekend
Membership plans — tier names, benefits, pricing
Emergency protocol — what qualifies, who's on call, how on-call dispatch works
Gas / CO2 safety protocol — your current script for gas-smell, CO alarms, electrical hazards
Warranty standards — what's covered under parts-only, labor-only, full warranty for new installations
Top 20–30 FAQs
Decisions you'll need to make
Voice — Dante (warm, direct, trades-appropriate) is the default
Disclosure — we recommend proactive disclosure
After-hours coverage — AI handles, or AI routes to after-hours call service?
Emergency dispatch rules — does the AI dispatch directly, or does it route to a human dispatcher first?
Service-area decline — do you partner with other companies for out-of-area referrals?
Payment collection — do you collect deposits or pay-at-call? (Enable PCI-compliant DTMF masking if yes)
Multi-trade routing — if you cover multiple trades, how the AI determines which to route to
Service area configuration
Upload your zip-code list, cities, or polygon map
Define any partial-coverage edge cases
Configure out-of-area responses and partner referrals (if any)
Set distance-based pricing tiers if you have them
Dispatch integration
ServiceTitan — API integration for booking, dispatch board visibility, technician ETA
Housecall Pro — API integration for bookings and customer records
Jobber — API integration for scheduling and customer records
FieldEdge, Service Fusion, RazorSync — API-based or direct database integration depending on setup
Simpler setups — Google Calendar or Outlook-based booking with message-based handoff to dispatch
Emergency protocol configuration
Gas emergency trigger words — "smell gas", "gas leak", "carbon monoxide", "CO alarm"
No-heat / no-cool triggers — weather-dependent prioritization (emergency in winter for heat, summer for cool)
Vulnerable-resident flags — elderly, infant, medical concerns trigger priority dispatch
Water-damage urgency — for plumbing-inclusive shops, flood / leak / sewage back-up get priority
Electrical hazards — for electrical-inclusive shops, sparks / smoke / shock get 911 guidance
Pricing and sales configuration
Diagnostic fee explanation — when it applies, when it's waived, when it's credited against repair
Estimate vs. repair distinction — estimates vs. billable service calls
Membership plan pitch — should the AI proactively offer the plan? Only on request? Never?
Upgrade estimate booking — if the caller wants a new install quote, who books the comfort advisor visit?
Call forwarding
Your current phone system — setup guides in our Call forwarding collection
Forwarding rules — always, after-hours, overflow
Emergency routing fallback — if the AI fails, where calls should go
Go-live sequence
Day 1: kickoff, requirements, service area mapping, FSM integration begins
Day 2–3: knowledge base, emergency protocols, pricing configuration
Day 4–5: internal test calls, your team tests real scenarios
Day 5–7: soft launch (after-hours first), then full rollout
Weeks 2–4: daily transcript review, tuning, especially for emergency handling
What to measure after go-live
Call containment rate
Dispatch accuracy (are bookings landing correctly on the board?)
Emergency response time
Service-area rejection rate (is the AI correctly identifying out-of-area calls?)
Lead capture (are new-customer calls becoming booked appointments?)
Technician ETA accuracy
Related articles
Start the process
Book a demo and our team will walk you through this checklist and configure a live HVAC Dispatch agent for your specific service area and trades.