Real conversation, not menus.
No press-1-for-X. Callers describe what they need; the receptionist understands and responds at human pace.
AI books real appointments during the call. Live calendar, drive-time aware, tech routing — confirmed before the caller hangs up.
When a caller says they want to schedule, the AI asks what service they need and their address. It then checks your calendar in real time, accounting for the time the job will take, the drive time from wherever the assigned tech's last appointment is, and the specific tech's schedule and skills. The AI offers two or three specific options ('I can get you Thursday at 10am or Friday at 2pm — which works better?'), the caller picks one, and the appointment is booked instantly. The calendar system — Google Calendar, Cal.com, or whatever you use — gets the new appointment within seconds. Both the customer and the assigned tech get a confirmation text with the details.
No press-1-for-X. Callers describe what they need; the receptionist understands and responds at human pace.
Emergency lines wake your on-call cell. Routine bookings drop into the calendar. Sales tire-kickers get triaged.
Full transcript, lead details, and confirmation hit your phone — so you can show up without re-asking.
Real scenarios we hear about most. Cards link to the matching trade page when one exists.
Caller asks for a tune-up; AI sees the assigned tech will be in the neighborhood Thursday morning, offers a drive-time-sensible slot at noon. Job booked before the caller hangs up.
See the HVAC →Electrical caller asks about a 200A panel; AI knows only your senior electrician handles those, offers only his slots, keeps the apprentice free for smaller service work.
See the Electrical →Recurring pest-control customer calls about a new ant issue; AI adds the free re-treatment to the next scheduled route stop in their neighborhood — no separate truck roll required.
Every service-trade shop has a version of the same broken loop: the caller asks when you can come out, your receptionist says 'let me check and call you back,' the receptionist forgets, and the caller books the next shop. Even when the callback happens, half the callers have moved on. The loop wastes time on both sides, costs you jobs, and exists only because your phone and your calendar don't talk to each other. RingDesk's Appointment Booking eliminates the loop. The AI has live access to your calendar — the same calendar your techs work from — and books the appointment during the call. It checks real availability, respects drive time between existing jobs, routes to the right tech based on the service requested, and sends a confirmation text to both the customer and your team before the call ends. The caller hangs up with a specific day and time, not a promise to be called back. Your tech gets the job on their schedule with no manual handoff.
The booking flow happens in the conversation without the caller noticing it as a separate step. After the AI has captured the basics — name, phone, address, service — it checks your calendar for available slots that match. The availability check is live; if another call booked Thursday at 10am five minutes ago, the AI won't offer that slot. The AI offers two or three options rather than a long list. Offering three options increases conversion dramatically over asking 'when would you like to come in' — callers don't know what you have, and making them ask one at a time exhausts both sides. Two to three specific times with a clear ask ('which one works for you?') closes bookings fast. When the caller picks a time, the AI confirms the details back ('so I've got you Thursday the 18th at 10am for an AC tune-up at 422 Oak Street, is that right?') and then saves the appointment. If the caller says 'actually make it 2pm,' the AI adjusts and re-checks availability for 2pm. The whole negotiation happens in natural conversation. If none of the offered times work, the AI can go back and check more availability — 'the week after, or would a Saturday work?' — without ending the call. If nothing works on your calendar, the AI captures the caller's preferences ('prefers afternoons, prefers Wednesdays') and promises a callback when a slot opens. You get a note on the lead explaining the mismatch so you know to reach out when space frees up.
Behind the scenes, the AI considers multiple constraints before offering a slot. First: is the slot open on the calendar? (Obvious, but critical — no double-bookings.) Second: can a tech actually get from their last job to this new job in time? Drive-time calculation uses real maps data, not rough estimates. If your tech's 9am job is across town from the proposed 10am appointment, the AI won't offer 10am — it'll offer 10:30 to account for the drive. Third: does the tech have the right skills? If the call is for a panel upgrade and only your senior electrician is certified for that work, the AI won't book the job with your apprentice. Skills are configured per-tech in your account, and the booking logic respects them. Fourth: is the slot inside the tech's working hours? The AI won't book a 6am appointment for a tech whose hours start at 8am. Holiday and time-off calendars are respected too — if a tech is on vacation next week, their slots don't appear. Fifth: is the job type compatible with the time of day? Some shops only do new-construction walkthroughs in the morning and service calls in the afternoon. You configure the rules, the AI respects them. All of this happens in under two seconds from the caller's point of view. They hear the AI pause for a half-beat and come back with specific options.
For any business where techs drive to jobsites, drive time is the silent killer of daily schedules. A shop can look like it has three open slots on Thursday but actually have zero slots that are sensible — because the three open windows are on three different sides of town and no tech can realistically make them all. RingDesk's Appointment Booking handles this. When the AI considers a slot, it calculates the drive time from the tech's previous appointment (or from their home base if it's the first job of the day) to the proposed address. Using real maps data, traffic-aware when possible, the AI adds a buffer that reflects actual driving conditions. If the buffer pushes the start time past the window of availability, the AI won't offer that slot. This prevents the worst failure mode of every service-trade booking system: the booked appointment the tech physically cannot reach on time. Those appointments become the customer calls that destroy your reviews — 'my appointment was at 10am and the guy didn't show up until 11:30.' With drive-time-aware booking, that stops happening because the slot would never have been offered in the first place. For shops where drive time isn't relevant — a dental practice where all appointments happen in one building — you can turn drive-time checking off in settings. The AI defaults to the faster mode (simple calendar check) and skips the maps lookups.
HVAC shop, spring tune-up season: A homeowner (Brian Hayes) calls Tuesday afternoon asking about a spring AC tune-up. The AI captures his address, checks the calendar, sees that one tech will be finishing a nearby job on Thursday at 11am and has a 1pm opening. The AI offers Thursday at 12pm (accounting for drive time + a 30-minute buffer) and Friday at 3pm. Brian picks Thursday. Appointment booked, tech gets the text, Brian gets the text. No callbacks, no checking with the office, no 'we'll get back to you.' Plumbing shop, Saturday morning emergency: A caller (Doug Reed) says he's got a leaking water heater in his basement. The AI tags it high-urgency, routes the booking logic to check for same-day availability, and finds that one tech is finishing a 10am job across town. The AI offers 'I can have someone out around 1pm today' — which accounts for drive time plus a buffer. Doug accepts. The tech gets the emergency-tagged text, which shows in his queue as the next job. He never has to call the office to confirm — it's already on his schedule. Electrical shop, panel upgrade routing: A homeowner (Paul Nelson) calls asking about a 100A to 200A panel upgrade. The AI knows (from flow configuration) that panel upgrades over 100A require a senior electrician, not the apprentice. It checks only the senior electrician's schedule, offers two slots next week, Paul picks one. The apprentice never sees the job on his route — the routing rule kept the work with the right tech. Pest control, multi-location route optimization: A recurring customer (Carl West) calls asking to add an extra treatment because of a new ant issue. The AI sees the caller is on the Wednesday route and has a quarterly treatment already scheduled in three weeks. Instead of booking a separate new appointment, the AI offers to add a free re-treatment to next Wednesday's existing route stop — because the tech will be in that neighborhood anyway. Carl accepts; the tech gets an updated job ticket for next Wednesday with a new note. Dental practice, recurring cleaning: A patient (Steve Graham) calls to schedule his six-month cleaning. The AI recognizes him as a returning patient, pulls up his preference for Tuesday mornings, checks the hygienist he normally sees, and offers the next two Tuesdays at 9am — both with his preferred hygienist. Steve picks one, done. No 'let me transfer you to scheduling,' no hold music, no callback.
When an appointment gets booked, the assigned tech gets a text message with the key details: caller name, phone, address, service, time, any notes the AI captured. The appointment also appears on the tech's schedule view — the same schedule they work from every day. Techs don't need to open the RingDesk dashboard to see new bookings; the schedule they already use is the source of truth. For shops with an existing scheduling tool (ServiceTitan, Housecall Pro, Jobber, etc.), RingDesk pushes the appointment to that tool so it appears in the schedule the tech already opens. This means the AI-booked job looks identical to a job booked manually by a dispatcher — no special workflow, no separate view, no retraining.
The AI books appointments for services and techs you have configured in your account. If a caller asks for something your shop doesn't offer — a service category not in your flow — the AI won't make up an appointment. It'll capture the request as a lead and flag it for your team to review. Drive-time calculations use maps data and are generally accurate within a few minutes. On a day with unusual traffic (big storm, event, construction), the buffer might be too tight. Most shops set their default drive buffer to add an extra 10 to 15 minutes of safety margin on top of the map estimate. If your calendar has conflicts the AI can't resolve automatically (for example, two jobs both partially inside the same time window due to a manual edit), the booking flow will flag the conflict on the lead and ask your team to review before confirming. The AI won't create a booking over the top of a conflict silently. For appointments requiring complex qualification before booking — legal intake conflict-of-interest checks, medical intake with pre-visit questionnaires, custom enterprise onboarding — the AI captures the lead and hands it to your team rather than booking directly. You can configure specific flow branches to skip direct booking in favor of lead capture. Cancellation and reschedule by the customer is supported via the confirmation text's reschedule link. If the caller wants to cancel by calling the AI directly, you can add that branch to your flow — by default, cancellations by phone are routed to your team to handle.
1. Connect your calendar in Settings → Integrations. Google Calendar and Cal.com are supported directly; other scheduling systems connect via webhook. If you use ServiceTitan, Jobber, or Housecall Pro for scheduling today, the integration pushes bookings there so your techs see them in their usual view. 2. Add your techs in Settings → Team, with their working hours, service skills, and home base address (for drive-time calculations). If a tech is senior for certain service types, mark them on the service-skills matrix. 3. Configure your services in Settings → Services. Each service type has a default duration, a price (if fixed), and a required skill level. The AI uses these to offer accurate time windows. 4. Set drive-time rules. Default is to add the maps-calculated drive time plus a 15-minute buffer. You can adjust globally or per-service (quick emergency calls might use a smaller buffer; careful installation jobs might use a larger one). 5. Turn on the booking node in your flow. Industry templates have it enabled by default; custom flows may need you to drag the booking node into the relevant branches. 6. Test. Call your test number, ask for an appointment, and verify the AI offers realistic times, books correctly, and your calendar receives the new appointment. If anything looks off, check the service configuration or the tech schedule.
The moment a caller hangs up without a booked appointment is the moment you start losing them to the next shop in their search results. Every step between 'caller wants to book' and 'appointment is on the calendar' is a chance to lose the sale. Callback-based booking turns a one-minute phone call into a one-day sales process — and a day is long enough for the caller to try someone else. Real-time booking collapses that entire process into the single conversation. The AI isn't just taking a message; it's closing the sale. That's the difference between a receptionist that captures leads and a receptionist that fills your calendar.
Google Calendar and Cal.com are directly supported. Other systems (ServiceTitan, Jobber, Housecall Pro, FieldEdge, Workiz, PestPac, Dentrix) are supported via webhook, so appointments appear in the tool your techs already use.
Yes. The calendar is always live. A caller at 11pm can book a morning appointment and your tech will see it on the schedule when they check it.
Yes. Every booking does a real-time calendar check immediately before confirming. If two callers try to book the same slot seconds apart, the first one wins and the second is offered an alternative.
Yes. In flow settings, you can mark specific service types as 'lead capture only' so the AI collects details but doesn't auto-book. Your team reviews and books manually. This is common for complex quotes or high-value jobs.
The confirmation text includes a reschedule link that opens a self-serve reschedule page. Customers can pick a new slot from the currently available windows. If they call back to reschedule instead, the AI handles it the same way the original booking.
It matches the service requested with the techs who have that skill, then picks the tech whose schedule and location make the most sense for the requested time. You can override routing rules per tech in Settings.
Mark them unavailable on their calendar (or use your existing time-off workflow). The AI won't offer their slots while they're out. When they return, their slots become bookable again.
The caller gets an automatic confirmation text right after booking, with the appointment details and a reschedule link. The assigned tech gets a separate text with the job information.
Yes. If the caller requests same-day service and your schedule permits, the AI will offer same-day slots. If nothing is available today, it explains the earliest possibility and offers to flag the lead as urgent for a possible squeeze-in by your team.
Every feature on this page rides the same six guardrails. Built into the product, audited on every call, fall back to your existing line if anything trips.
Every price comes from the table you upload during setup. Off-book questions become "I'll have the owner call you back," not a guess.
Zero-hallucination quote engineYou write the keyword list (burst, no heat, gas, smoke, sparking). The AI escalates the moment one fires — no menu, no hold, no triage delay.
Median escalation: 4.2 secConstrained to your live calendar. If a slot fills mid-call, the next available one surfaces immediately. No phantom appointments.
Source-of-truth: your calendarTone, profanity, and "get me a human" all trigger an instant transfer. No "let me help you first" loop. The customer gets you, fast.
Tone-detection failoverCarrier failover forwards calls to your existing number and raises an ops alert. The claim is simple: fail closed to a human line, not to silence.
Forwarding fallback · ops alertListen to any call in the dashboard within minutes. Catch a misquote, retrain the flow, refund a customer — all from one screen.
Recording + transcript on every callEvery feature live on day one. Starter, Growth, Pro — same platform, same receptionist.
Live the same morning. Cancel any time.
Start trialDrop in your call volume and average ticket — see your weekly leak in twenty seconds.
Open calculatorA short field guide on the four moves shops use to lock in 99% pickup.
Read the guide