Product Webhooks & APIPro plan

Push it into whatever you already run.

Point an endpoint at your workspace and conversation events arrive there, signed, filtered and retried. Or read the same data through a small REST API. Zapier sits in the middle of both when you would rather not write anything.

  1. 1Something happens in the inbox
  2. 2We sign the event and POST it to your URL
  3. 3A failed delivery retries, then tells you
  4. 4Or poll the REST API for the same data

Why it exists

Every business has one system the inbox has to reach

It is a CRM for one team, a spreadsheet for another, an internal tool nobody outside the company has heard of for a third. No connector list is ever going to cover that, and pretending otherwise is how integrations pages end up listing forty logos and solving nobody’s actual problem.

A signed webhook and a small API cover all of it, because a URL is a thing every system has. What matters is whether the plumbing is honest: that a failed delivery is visible rather than silently dropped, that you can prove a request came from us, and that the thing on the other end cannot be used to read what the app itself would refuse.

What you get

In the product today.

Thirteen events, eight on by default

New conversation, label applied, contact updated, deals created, updated and deleted, tickets created and updated — plus every message, delivery status, conversation changes and follow-ups, off by default.

“Every message” is off on purpose

A workspace doing two thousand messages a day would spend sixty thousand Zapier tasks a month on that one event. It is there if you want it, and it is a decision you make knowingly rather than a default that quietly bills you.

Signed, so you can prove it was us

Every delivery carries a timestamp and an HMAC-SHA256 signature over the body. Your receiver can verify it rather than trusting whatever arrives at a public URL.

A delivery log, not a shrug

Every attempt is recorded with its status and response code. Redeliver one by hand, or fire a test at a new endpoint before you rely on it.

Broken endpoints turn themselves off

Repeated failures, or a 410 Gone, disable the endpoint and record why. A dead URL stops being retried for ever, and you are told which one it was.

A REST API, small on purpose

Read chats, messages, contacts, deals, tickets and labels; send a message, apply a label, write a note, set a follow-up. It uses the same tokens as the AI-agent connection and the same visibility rules as the inbox — there is no mode where the API sees everything.

When you would use it

Situations this is for

  • A new conversation creating a row in your own system automatically.
  • A Zap that posts every new enquiry into a Slack channel.
  • A ticket status change updating a dashboard you built yourself.
  • Reading yesterday’s conversations into a warehouse on a schedule.

Questions

Webhooks & API — questions

Does this work with Zapier?

Yes, from both ends. A Zapier Catch Hook is a URL, so it is a webhook destination like any other, and the REST API exists specifically so a Zapier app can poll and act. Make or n8n work the same way.

What about Salesforce, Zoho or monday.com?

Not as native connectors — those go through a webhook or a Zapier step. HubSpot and Pipedrive are the two we connect to directly.

Can a webhook leak a private conversation?

No. An endpoint has no user attached to it, so it is given the one visibility tier defined without a person in it — workspace scope, which excludes private chats entirely. It is structural, not a filter that a future change could forget.

Can I point one at an internal address?

No, and the check runs when the message is sent rather than only when the URL is saved — a save-time check is decorative against DNS rebinding. Destinations resolving into private ranges are refused.

What happens if my endpoint is down?

The delivery is retried. If failures keep stacking up, the endpoint is disabled with the reason recorded, so it fails loudly once rather than quietly for a month.

Which plan is this on?

Pro, and owner or admin only. An endpoint forwards conversations at workspace scope, so letting an ordinary member create one would be a way to read, through a third party, what the app itself refuses them.

Stop running your best channel off a phone in a drawer.

Connect your first number in five minutes. Keep the number, keep the history, and finally see the whole thing.

No credit card required.

Chat on WhatsApp