MapleReceptionist News - what we have shipped

MapleReceptionist now tells you where a caller is phoning from and how likely the call is to be junk - before you ever pick up. Every inbound call is enriched with the caller’s city and region, and scored for scam or spam confidence, with the result written to the call record, the per-call email summary and the in-dashboard call log. The business problem is wasted time and missed real customers. A two-person garage that stops work for every robocall about “your vehicle’s extended warranty” loses billable hours; a clinic that screens too aggressively risks hanging up on a real patient. Caller-city plus scam confidence lets the AI handle the obvious junk quietly while flagging the genuine new-customer call from the next town over as worth a callback. Owners get the context at a glance in the summary email - city, likely intent, and a junk-confidence read - so a returned call is a returned call to someone who actually wants to book.

Following live-call review, MapleReceptionist’s warm-transfer flow was rebuilt so the AI reliably recognizes when a caller wants a human - “can I talk to the owner,” “is Josh there,” “this is an emergency” - and bridges the call to the right extension with the correct caller ID attached. For a small business, a transfer that fails is a customer lost. An emergency tow caller who gets stuck in a loop hangs up and dials a competitor; a hot lead who asks for the owner and instead gets a voicemail rarely calls back. The updated flow resolves the destination extension, places the outbound leg, and bridges only when a human answers - with a graceful fallback to a message or callback if no one picks up. The result: the calls that most need a person actually reach one, and the owner’s real phone number shows up correctly on the bridged call.

MapleReceptionist now handles business hours, after-hours and holiday rules correctly across every phone number on an account. Each line can carry its own open hours, its own after-hours behaviour and its own holiday calendar, so a multi-location shop or a clinic with separate booking and billing lines routes each number the way it should. Missed calls cluster at exactly the wrong times - lunch, after 5pm, weekends and stat holidays - which is when most SMBs are closed and most US-built answering services get the timezone wrong anyway. Atlantic-Time-aware hours mean “after hours” means after your hours, not a server’s. The AI answers, books or takes a message according to the rule that applies to that specific number at that specific moment - so a caller on Labour Day gets a proper holiday greeting and a booking option, not a dead ring.

Every call MapleReceptionist answers now generates a per-call email summary sent straight to the business owner: who called, what they wanted, what the AI did, and any booking, message or callback that came out of it. For an owner who is under a truck, in a patient’s mouth, or on a job site, this is the difference between knowing and guessing. Instead of a voicemail light and a mystery, the owner gets a readable recap they can act on in ten seconds - call the lead back, confirm the booking, or ignore the obvious junk. The summaries were hardened during a live garage rollout so they read like a note from a competent front-desk person, not a raw transcript. Nothing falls through the cracks just because the phone rang at a bad time.

MapleReceptionist now recognizes inbound callers by their phone number and pulls in what it already knows about them - so a returning customer, tenant or patient is greeted in context instead of starting from zero every time. Caller-ID is matched against the account’s contacts and lead records, with role precedence so the AI treats staff, owners, vendors and customers appropriately. Small businesses live on repeat relationships. Making a loyal customer re-explain who they are on every call feels like a chain call-centre, not the local shop they chose. With caller recognition, the AI can say “hi again,” route a known vendor differently from a new prospect, and skip questions it already has answers to - faster calls, warmer service, and fewer details to re-collect.

MapleReceptionist answers in English and Canadian French - including Quebec French and Maritime “Frenglish” patterns - on every plan, with no language add-on or upgrade tier. The signup and client surfaces are fully bilingual too, so a francophone owner runs the whole product in French. In Atlantic Canada and Quebec, a receptionist that cannot take a French call is not a receptionist. A caller from Shediac or Dieppe who reaches an English-only US answering service simply hangs up - and that lost call is a lost customer in a market where bilingual service is the baseline expectation. Because both languages ship standard, a business never has to choose which half of its callers to serve well. The AI detects and responds in the caller’s language automatically.

MapleReceptionist gained two-way SMS with a from-line picker plus a built-in browser softphone, so a business can text and call back from its own business number - right from the client portal, no second cell phone required. Most missed-call revenue is lost in the gap after the call: the lead who left a message and never heard back, the booking that needed one quick confirmation. Texting from the business line, not a personal mobile, keeps that follow-up professional and inside one record. Combined with the AI’s call handling, a caller who could not be served live gets a fast text follow-up from the same number they dialled - closing the loop while the lead is still warm.

MapleReceptionist books appointments directly into the connected calendar during the call. The booking questions are client-configurable - a dental clinic, a law office and an HVAC shop each collect what they need - and the AI confirms the caller’s details, including a phone-number confirmation, before it writes the appointment. A graceful fallback takes a message if no calendar is connected. For an SMB, a booking taken live is worth far more than a voicemail to triage later. The caller who reaches voicemail at 7pm books with whoever answers next; the caller who books on the spot is on the schedule. Moving booking logic into the AI - rather than a rigid phone tree - means callers can describe what they want in plain language and walk away with a real, confirmed time slot.

MapleReceptionist shipped true live call transfers to staff extensions: the AI resolves the target extension’s real phone number, places parallel outbound calls, bridges the caller through the moment a human answers, and handles a no-answer gracefully rather than dropping the caller. A receptionist that can only take messages is half a receptionist. Sometimes the caller needs a person now - an urgent service call, a decision only the owner can make - and the ability to actually hand the call off is what makes the AI a front desk instead of an answering machine. Bridge-on-answer means the caller is not transferred into a void: if the extension does not pick up, the AI stays on the line and takes care of them.

MapleReceptionist launched as a 24/7 AI phone receptionist built and hosted in Canada for small and medium businesses - clinics, law firms, garages, HVAC contractors, property managers and home services. It answers every call, day or night, in English or French, with call data kept inside Canada and PIPEDA plus PHIPA compliance built in. The problem it was built for is simple and expensive: phones that ring into voicemail. A missed call is a missed customer, and for most SMBs those calls land after 5pm, on weekends, or while the one person who answers the phone is already busy with another customer. MapleReceptionist answers on the first ring, every time, and is priced transparently in CAD on a month-to-month basis - so a two-truck shop and a four-chair clinic can both afford to stop losing calls.