You have automated the inbox, the forms and the invoices. The phone is the last job still run on memory and sticky notes. Talkify is the Make.com AI receptionist that answers every call in a real New Zealand voice and then posts the whole call to your scenario as tidy JSON, so Make phone call automation starts the second the caller hangs up. Scenarios you built back when Make was Integromat work just the same.
Put your name down and we will point your first call at your Make scenario and walk you through it. No setup fee, no lock-in, and your phone number stays exactly where it is.
One call out as data and a way back in. Everything after that is whatever you drag onto Make's canvas.
Paste your Make webhook address in as a destination and every finished call is sent there as JSON, with nothing to poll and no email to parse.
Caller name, number, email and company, the summary, each captured detail as its own field, the outcome, the length and the time the phone rang.
A destination can take only one agent's calls, only calls over a set length, only after-hours calls or only certain outcomes, so your scenario runs less often.
Posts go to https addresses only. Add your own headers if the far end wants a key, and set a secret to have each body signed with SHA-256.
An HTTP module calling our API starts an outgoing call with a brief and your questions, and the answers come back to Make as data.
Post a lead from any app Make connects to and it lands in your Talkify leads, where your follow-up automations ring or text them.
A missed call never becomes a bundle. The job that rang at 7:42 on a Saturday and hit voicemail does not exist anywhere Make can see. No row, no alert, no follow-up, just a caller who tried the next name on the list.
A voicemail is unstructured work. Somebody still has to listen, type it up and decide where it goes. That is the manual step your automations were meant to remove.
A call answered by Talkify arrives ready to route. Named fields, a clear outcome and a transcript behind it, posted within seconds, at any hour, however many people ring at once.
What happens between the tenant saying hello and the plumber being booked, with no person in the loop until one is needed.
Your greeting, your business name, a local voice, first ring. It handles as many calls at once as come in, so a Monday morning rush is never a queue.
The details you asked it to collect: unit number, job type, preferred time, whatever your process needs. Each one becomes its own field, not a paragraph.
A plain summary is written and the call is given an outcome, from a happy caller to one waiting on a call back. That is the field your Router branches on.
Each destination whose filters match gets its own post. A slow or broken one never holds up the others, and a failed delivery shows in red on your screen.
Add a row, open a job, message the on-call manager, update a CRM. By default Make runs the scenario the moment the data arrives.
If the scenario decides someone needs a follow-up call, one HTTP request to Talkify starts it, and a second webhook brings the answers home.
This is the shape of the body your Custom webhook receives for the unit 4 call. Press Send test in Talkify and Make learns it in one go, so every module after the trigger can map the caller, the outcome and each detail the agent took down. Want us to look at your scenario with you first? We are happy to.
{
"event": "call.completed",
"agent": "Harbour Property after hours",
"caller": {
"name": "Aroha Ngata",
"number": "+64 21 555 0107"
},
"summary": "Hot water cylinder leaking in unit 4. Switched off at the board.",
"captured_details": { "unit": "4", "urgency": "urgent" },
"outcome": "callback_requested",
"duration_secs": 94,
"called_at": "2026-09-26T07:42:08+12:00",
"transcript_url": "https://…"
}
Make turns up wherever one person has quietly become the office's integrations department. These are the businesses that ring us.
Maintenance calls at all hours, each one needing a job raised and the right tradie told. AI reception for real estate.
Faults · tenant callsFirms whose quoting and scheduling already run through scenarios, and whose phone did not. AI reception for trades.
Quotes · call-outsTeams that route every enquiry to the right person automatically and want calls routed the same way. AI reception for professional services.
Enquiries · triageMost phone answering stops at telling a person. Structured data is what lets the next step happen without one.
Because the outcome and each captured detail arrive as separate fields, a Router filter can say "outcome is callback_requested and issue contains leak" instead of text matching an email body that changes shape every time. Build the outgoing half with the payload builder.
Make is at its best when a single event has to go three ways. A new enquiry, an unhappy customer and a booking change can each take their own route from one webhook, which keeps the logic in one place you can read and fix.
Most AI answering services sold in New Zealand are overseas platforms with a local logo. Talkify runs on our own local SIP network, keeps data in New Zealand and Australian data centres, and answers in voices your callers actually recognise. That changes how every call sounds, and where every recording lives.
Tell us which system you run and how your calls come in. A real person at Talkify replies, and we set the connection up with you.
Flat NZD pricing, published on the site. No setup fee. No lock-in. Built in Ōtautahi Christchurch and operated by Talkify Limited.
We run our own local SIP carrier network, so your calls connect close to home. Less lag, no talking over each other, and the clear, close sound of a local call.
A full range of Kiwi and Aussie voices, not an American accent doing its best. Callers relax when the voice sounds like them, and relaxed callers book.
Regional phone numbers across New Zealand and Australia, so a Christchurch caller rings 03 and an Auckland caller rings 09. Local numbers get answered; unknown ones get screened.
For data sovereignty, everything is stored in New Zealand and Australian data centres: recordings, transcripts and customer details held close to home under local privacy law.
Full outbound service, not just answering. Your agents confirm bookings, chase missing details and run marketing call campaigns, with every conversation transcribed.
Load a contact list and the agent rings through it, asks your questions, and hands back transcripts with the answers extracted and analysed. Offers, follow-ups or a straight survey, at a scale no clipboard reaches.
One agent or a whole fleet answering unlimited simultaneous calls. Overflow, after-hours or the entire front line, with warm transfer to a human whenever it matters.
Calendars, CRMs like HubSpot and GoHighLevel, gym platforms, WhatsApp and SMS, plus Zapier and a full REST API when you want to wire up something custom.
No DIY prompt wrangling. We build your agent around your business, brief it on your services and pricing, tune it with you, and keep tuning after you're live.
Native apps for iPhone and Android push you each call summary the moment it ends: who rang, what they wanted, what the agent did about it.
Full transcripts, sentiment, outcomes and intent on every call, rolled into dashboards that show what your callers actually want, not just how many rang.
The questions we hear most. If yours isn't here, talk to us and a person will answer it.
The feature doing the posting: Slack, Teams, Make or your own endpoint, each with its own filters.
The same call data into n8n workflows, including an n8n you host yourself.
What the agent does when a scenario asks it to ring someone: reminders, follow-ups and quick surveys.
Where a lead posted from Make ends up, and the automations that ring or text it back.
On HubSpot? The direct connection files the caller and the transcript with no scenario in between.
Everything Talkify connects to, and what your agent can do once it is connected.
Or run it from your pocket