Setup checklist for an insurance AI receptionist
A complete checklist for deploying the 365agents Insurance Agency agent — AMS integration, carrier protocols, licensed-advice boundaries, claims FNOL workflow, and producer routing rules.
Written By Catherine Weir
Last updated 9 days ago
Deploying an AI receptionist for an insurance agency takes 1–2 weeks of elapsed time. The regulatory boundary (no interpretation of coverage, no quoting, no binding) is straightforward to configure once you identify the specific call types you want handled by the AI vs. routed to a licensed producer.
Before the kickoff call
Agency management system (AMS) — EZLynx, Applied Epic, AMS360, NowCerts, HawkSoft, QQ Catalyst, or similar (with admin credentials)
Lines of business — personal auto, home, life, umbrella, commercial auto, commercial property, workers' comp, etc.
Producer roster — producers with their lines of business, new-client availability, territories
Carrier list — which carriers you represent and any carrier-specific claim protocols
Claims handling protocol — direct-to-carrier, direct-to-adjuster, or through a CSR; specific carriers' FNOL processes
Service request workflows — how ID cards, COIs, and address changes get processed today
Payment handling — what payments come to the agency vs. carrier portals; PCI DSS SAQ D posture
Licensed-advice boundary — the specific topics you want the AI to route to a producer rather than address
Renewal workflow — who handles renewal calls, how policy changes are processed
Top 20–30 FAQs
Decisions you'll need to make
Voice — Atlas (calm, trustworthy) is our default for financial contexts
Disclosure — we recommend proactive disclosure
After-hours handling — AI all hours (claims FNOL benefits from 24/7) or only during business hours
Quote intake depth — how much to collect before routing to producer
Direct payment collection — does the AI process payments (requires PCI DSS SAQ D configuration) or always route to a portal
Certificate turnaround — what timeline to promise callers for COI delivery
Line-specific routing — personal auto producer vs. commercial producer vs. life producer
Licensed-advice boundary configuration
The AI should be explicitly configured to:
Never quote coverage, rates, or premiums
Never interpret policy coverage ("is this covered?")
Never bind coverage or confirm coverage is in place
Never recommend specific products or carriers
Always route these to a licensed producer
Answer factual knowledge-base questions about the agency itself (who we are, which carriers we represent, what lines we write) freely
AMS integration
EZLynx, Applied Epic, AMS360 — API-based integration for customer lookup, policy data, and activity logging
NowCerts, HawkSoft, QQ Catalyst — API or Zapier-based integration
AgencyZoom — integration for lead intake workflow
Proprietary systems — most AMSes have at least read-only API access; we'll work with yours
Claims FNOL configuration
Specific FNOL question sets by line (auto, property, liability, etc.)
Carrier-specific claim numbers and protocols
Escalation thresholds (injury involved, catastrophic damage, etc.)
Adjuster routing rules
Sub-producer / CSR handoff paths
Payment configuration (if applicable)
PCI DSS SAQ D service provider agreement with 365agents
DTMF masking enabled for phone payments
Payment confirmation email/text templates
Tokenized payment link option for callers who prefer
Call forwarding
Current phone system — setup guides in our Call forwarding collection
Forwarding rules — ideally always-forward for operational efficiency
Claim / adjuster after-hours routing
Go-live sequence
Week 1: kickoff, requirements, AMS integration, producer routing setup
Week 2: knowledge base, FNOL protocols, licensed-advice boundary configuration, test calls
Week 2: soft launch (overflow or after-hours first), monitor transcripts daily
Week 3+: full rollout, refinement based on call patterns
What to measure after go-live
Containment rate by call type (ID cards and COIs should be very high)
Producer-time reclaimed (how many calls no longer require a producer?)
Claims FNOL completeness (is the agency getting clean intakes?)
Quote lead quality (are the handoffs to producers useful?)
Licensed-advice boundary adherence (sample transcripts to confirm the AI isn't straying into interpretation)
Related articles
Start the process
Book a demo and our team will walk through this checklist for your specific lines of business and AMS. Most independent agencies are fully live within 1–2 weeks of kickoff.