Two minutes of conversation. Every qualifying question answered. Recording and transcript stored on the contact.
CareRM picks up the phone, asks the qualifying questions, books the consult on your live calendar, and hands off to a person the moment anyone asks. On the phone number you already have. You don't replace your phone system. We layer onto it.
The live demo includes a sample call you can listen to end-to-end. No phone setup required.
The daughter calling from the hospital hallway at 7 PM isn't going to leave a voicemail and wait for Monday. She's calling the next three agencies on the list. Whoever picks up first wins. The math hasn't changed in a decade. The only thing that's changed is what's technically possible at 7 PM.
of home service inquiries arrive after standard business hours.
of callers who hit voicemail never call that business back.
more likely to qualify a lead when you respond within the first hour vs. waiting 24.
Industry research: Invoca; Housecall Pro; Harvard Business Review (Oldroyd / MIT).
The usual workaround
Hire an answering service. Pay roughly a dollar or two per minute. Get a script reader who can take a name and a callback number, and not much else. And the second a caller mentions a diagnosis or a medication, you have a HIPAA problem, because most answering services don't sign a BAA.
CareRM picks up in two rings, runs the same qualifying conversation your best intake coordinator would, and books the consult on the calendar. With a HIPAA disclosure on every call and a BAA on file with every vendor in the path.
One inbound call. Seven things happen between the first ring and the recap text. You can listen to the recording, read the transcript, and see every decision in the audit log.
Always-on, after-hours only, overflow, or only on the lines you choose. Opens with the HIPAA disclosure required by your state.
Who the care is for, what they need, where they live, when they need to start, who's making the decision. The same questions your best coordinator would ask.
Before any follow-up text or callback gets queued, the caller is asked and the consent is logged with timestamp and the exact language used.
Reads the live calendar. Holds the slot. Confirms it on the call. Sends the calendar invite by SMS and email.
If the caller asks for a human, sounds upset, or hits a topic that needs an actual person, it warm-transfers to the on-call number. If nobody picks up in 8 seconds, it schedules a callback.
The caller gets a short SMS summarizing what was discussed, when the consult is, and how to reschedule. Quiet hours respected.
Recording, transcript, structured fields, consent event, and disclosure event all land in the contact record. Available to export.
Daughter calling from a hospital hallway about her mom. This is what your team wakes up to Wednesday morning.
Two minutes of conversation. Every qualifying question answered. Recording and transcript stored on the contact.
Consult booked on the live calendar mid-call. Confirmation sent before she hung up.
Recap SMS within 30 seconds. TCPA consent and quiet hours checked first.
CareRM is not a replacement for your phone system. It sits in front of it (or behind it) and answers the calls you want it to answer. VoIP, cell, desk phone, landline. You keep your existing number. Calls route exactly where you tell them to.
AI answers every inbound call, business hours and after.
Your team takes business hours. AI takes nights, weekends, holidays.
AI picks up only when your team doesn't answer within N rings.
Run AI on the intake line. Leave the office line untouched.
The moment a caller says a medication, a diagnosis, or a hospital name, the call contains PHI. Most AI voice vendors don't sign a BAA, log to everywhere, and route transcripts through systems you can't audit. That's a HIPAA exposure on every call. CareRM was built the other way around.
Each AI call opens with the disclosure your state requires (CA SB 1001, plus FCC ruling treating AI voices as artificial under TCPA). Per-state floor. Not tenant-configurable below it. Logged.
Every SMS, voice callback, and email checks consent_records and quiet hours before the network call. No consent, no send. No exceptions.
Voice runs on Vapi (BAA tier) over Twilio Programmable Voice. SMS on Twilio. Email on Postmark. BAA must be on file in vendor_baas before any vendor sees PHI.
PHI never lands in application logs, traces, or Sentry. Recordings and transcripts live in tenant-scoped storage with row-level security enforced at the database.
Voice and chat are included in the $497 per location flat rate. Carrier usage is pass-through at cost.
Part of the $497 per location, per month plan. No per-seat fee. No tiered upsell.
Twilio for voice and SMS. Postmark for email. Billed at carrier cost. We do not mark it up.
Total pass-through for a normal-volume location. We show you actuals before you commit.
The math is simple. Twilio voice runs roughly a penny or two per minute. SMS runs about a cent per segment. A typical agency does a few hundred minutes and a couple thousand texts a month. The total lands in coffee-budget territory.
Every AI-answered call opens with a short disclosure: it identifies the call as an AI assistant for your agency and notes that the call is recorded. The exact wording follows the state your location operates in (California SB 1001 has specific language, several other states are catching up, and the FCC's February 2024 ruling treats AI voices as artificial or prerecorded under TCPA). The disclosure is a floor, not tenant-configurable below it, and each one writes a row to the 45 CFR §164.528 accounting log.
Yes. If the caller asks for a human, sounds frustrated, gets into a topic that needs an actual person, or hits an escalation flag you defined, the AI warm-transfers to the on-call number. If nobody picks up within 8 seconds, it schedules a callback and sends a text confirming the time. You decide the on-call rotation per location.
No. It sounds like a person. We use the latest voice models with sub-300ms turn latency, so it can interrupt and be interrupted the way humans do. Most callers don't ask if it's AI. We tell them anyway because we have to.
Whatever was captured before the hangup is saved. If the caller provided a phone number or you can resolve their number to a contact, they get added to the pipeline with a 'call dropped' tag and the partial transcript. A coordinator can call back during business hours. Quiet hours are respected for any automated follow-up.
It's not a script tree. It's a real conversation backed by an LLM that's been trained on home care intake conversations and given your agency's specifics (services, geographies, pricing posture, things you do not serve). It handles digressions, callbacks to earlier turns, and the family member putting the phone on speaker mid-call. Inside the conversation it's pulling structured fields (care needs, location, decision-maker, urgency) that land on the contact record.
Yes. Greeting, voice, hours, qualifying questions, escalation rules, calendar, and on-call number are all per location. A 4-location franchisee can run four different greetings and four different on-call rotations from one workspace. Brand-wide policies (disclosure language, consent capture, quiet hours floor) are enforced everywhere.
The AI Voice + Chat product itself is included in the $497 per location, per month plan. Twilio and Postmark usage (voice minutes, SMS segments, email sends) are pass-through at cost. We don't mark them up. Most agencies see $20 to $50 a month of pass-through. We show you the actuals on a dashboard so there are no surprises.