Skip to main content
A lead fills in a form at 9:04. At 9:05 they are on the phone with an agent that knows what they asked for, books them into the client’s calendar, and writes the outcome back onto the contact, in GoHighLevel or here. That is what Team Follow Up does, in every sub-account your agency runs. The dashboard does all of it. The API does all of it too, so it fits into how your agency already works. A lead fills in a form at 9:04 and gets the campaign's trigger tag in GoHighLevel. At 9:05 the agent is on the phone with them and books them, and the result is saved back onto the contact. A lead fills in a form at 9:04 and gets the campaign's trigger tag in GoHighLevel. At 9:05 the agent is on the phone with them and books them, and the result is saved back onto the contact.

​
Start here

Quickstart

Create an agent, give it a number, hear it call you, switch it live. Five requests.

How it works

Agent, campaign, project and number, and how a lead moves through them.

Authentication

Create an API key, pick the narrowest scopes, send it as a bearer token.

Conventions

Ids, envelopes, dates, pagination, error codes and rate limits, once.

​
What do you want to do?

Call every new lead within a minute

Speed to lead: the trigger tag lands, the agent dials. Optional delay so a text lands first.

Work a lead over days, not once

Cadences: call day one, text day two, call again day four, tag and stop.

Only call when the business is open

Calling windows in the sub-account’s timezone, or the lead’s own.

Book, transfer or tag on the call

Capabilities decide what the agent may do and which outcomes a call can end in.

Recover leads you already have

Dry-run, then enrol a thousand tagged contacts through the same entry new leads use.

Prove what happened to every lead

Transcript, recording, outcome and cost on every call record. Metrics you can explain.

Do the CRM work after each call

One branch per outcome: tag, move stage, note, text, book.

Find out which opener books more

Two to five variants of one agent, leads assigned by weight, results per variant.

Charge clients for the calls you run

Rebill a sub-account at your own rate, take the payment from your client’s card, and see where every agency stands.

Run one agent across every client

Build the campaign once. Each sub-account’s calendar and pipeline resolve for it.

Let an AI assistant run it

Every published endpoint as an MCP tool except the ones that charge a card, filtered by the credential’s scopes.

​
How the pieces fit

  • An agent knows what to say: prompt, voice, skills, and what to record. See What an agent is.
  • A campaign decides who the agent calls and when: trigger tags, dispatch mode, calling window, caller ID. See What a campaign is.
  • A project is one sub-account. It keeps its contacts in GoHighLevel or in Team Follow Up, and its agents, campaigns, contacts and calls all live inside one. See Agencies, projects and sub-accounts.
  • A wallet pays for the calls. Your agency’s buys the minutes; a sub-account can have its own when you rebill it, funded by your client’s card. See Balances, spend and sub-account wallets.
  • Every term used in these docs is defined in the Glossary.

​
The API

Everything above is reachable over REST. The API reference is generated from the same route table the API serves, grouped by module, and each group links back to the guide that explains it. If you would rather talk than type, connect an AI assistant.