Quote, serve, and save — every call, every hour.

It answers in your branded greeting, captures the request, and either handles it directly or texts you depending on what you've authorized. Standard self-serve items — sending dec pages, emailing ID cards, taking billing questions, accepting COI requests, capturing add/remove vehicle or driver requests, updating mailing addresses — get logged and queued for your next business morning, with the client texted a confirmation. Quote intakes get a same-evening text with your quote form link. Claims and "I want to cancel" calls escalate to you in real time.

Yes. During onboarding you load your carrier list with each carrier's claims phone number. When a client calls for a claim, the AI asks which carrier the policy is with, confirms, and either reads the number out + texts it, or warm-transfers to the carrier's claims line directly — your choice per carrier. If the client uses an emergency trigger word you've defined (fire, injury, total loss, etc.), it skips that flow and dials you or the carrier's emergency line first.

It captures the lead and pushes it into your quote form (we support any URL-based form — your AMS quote intake, EZLynx, a Google Form, whatever). It does not bind policies — binding requires producer judgment, underwriting, and your e-sig flow, which is your job. The AI's job is making sure a real human-quality intake lands in your inbox at 6am Monday from a call that came in at 11pm Saturday.

Lead and servicing-request data lands in the portal immediately and can be pushed to your AMS via the platform integration we build per customer. EZLynx and Vertafore are the most-requested today and quoted as a separate setup project — they're real engineering work, not a checkbox. If you're on HawkSoft, AMS360, or QQCatalyst, we'll scope it during the demo. If you don't want an integration, the daily digest email + portal export covers it.

By default, the AI immediately texts and calls you so you can attempt a save while the client is still on the line — most agents tell us they save 30-50% of cancellation calls that reach a producer in time. You can change the trigger words (cancel, switch, shopping, GEICO, Progressive, anything else), change the escalation channel (text-only, call-only, or both), or have the AI itself attempt a soft retention pitch first if you'd rather. Whatever you pick, it's logged so you can review and tune.

Yes — English and Spanish are both standard. The AI detects the language from the caller's first words and continues in whichever they chose, including reading dec page details, ID card content, and carrier claims numbers in Spanish. The same flow is available for any other language during onboarding (Vietnamese, Tagalog, Portuguese, etc.) — each additional language is a small add-on for the prompt engineering and QA.

Every call gets a recording-disclosure played at the start (configurable per state to match two-party-consent rules in California, Florida, etc.). Recordings are stored for the standard 7-year insurance retention by default and are exportable. The AI is configured with your licensed states so it won't promise quotes outside your appointed footprint — it tells the caller you're not licensed there and offers to take a referral. TCPA SMS consent is captured before any outbound text. If your state DOI has specific disclaimers required, we add them to the opening greeting.

Yes. The service model is a tenant-config toggle — change it in your /client/settings/insurance panel anytime. Existing call history isn't affected; only the routing of future claim calls changes. Many agencies start Offloaded for the speed-of-setup and turn FNOL Capture on after 60-90 days once they have a feel for what the AI handles well.

Yes — and it's one of the highest-leverage problems we solve. The AI uses your producer-schedule + line-of-business + carrier-expertise + language matrix to route every caller to the right producer, even on a single shared number. You configure once during onboarding; it adapts as schedules change. The fallback (no producer free) is structured intake capture + queue with auto-page.

Probably not — Network tier is specifically for brokerages where the SUB-AGENTS are CAPTIVE under your brand (W-2 or fully-captive 1099, exclusive to your shop). If your producers are independent contractors with their own books — even though they work under your storefront brand — they should each be their own Solo Producer tenant on aiapps365. The brand affiliation is external to us. We'll walk you through the structure on a 20-min call.

Honestly: all are "scoped per customer" right now. We integrate during onboarding — typical 2-6 week build for a first-of-its-kind connector, faster on subsequent customers. If you're on a less-common AMS (Newton, Sagitta), we can still scope it. The integration goes through a generic event bus, so adding a new AMS doesn't require touching the core platform. We'll quote the integration cost upfront before contract.

First 20 customers per vertical (Insurance / HVAC / Plumbing — 60 total spots) lock their monthly base + add-on monthly fees for 24 months from signature. Per-minute voice and per-SMS overage rates ALSO lock for founding-cohort customers (non-founding customers' overage rates float with 30-day notice per MSA). Setup fees waived. To claim a spot you book a demo and sign before the 20th customer in your vertical signs.

Two things kick in. (1) The AI scales horizontally — there's no per-tenant call cap during a disaster surge. Whether you get 12 calls/day or 1,200 calls/day, every caller still gets answered in under 2 seconds. (2) Your emergency-bypass trigger words (active threat, injury, evacuation, total loss) route those callers straight to your on-call producer or carrier emergency line, bypassing the normal flow entirely. We'll review your disaster-mode config separately during onboarding.