Get a Scope
CRM

Close CRM AI automation and integrations

The short answer

Close is a sales CRM built around calling, email and SMS, which makes it a strong base for AI call summaries, coaching and follow-up drafting. Its REST API uses API keys over HTTP Basic auth or OAuth 2.0, webhook subscriptions are created through the API, and rate limits apply per endpoint group. Zapier has a Close app; n8n connects through its HTTP Request node because there is no built-in Close node.

What you can automate with Close

Call summaries onto the lead

Pull the call recording or transcript after each call, write a structured note, update the opportunity, and add follow-up tasks.

Read the build guide
Call coaching notes

Score calls against your talk track and objection handling and send each rep a short private note with one thing to improve.

Read the build guide
Follow-up drafts from the conversation

Draft the follow-up email or SMS from what was said on the call, saved for the rep to edit and send from Close.

Read the build guide
Inbound lead qualification

Score new leads from forms or ads, fill custom fields with fit and reasoning, and add qualified ones to the right Smart View.

Read the build guide
Lead enrichment

Fill company size, industry and contact roles on new leads from enrichment sources before the first call.

Read the build guide
Weekly pipeline brief

Summarize opportunity movement and activity levels by rep into a short written update for the sales lead.

Read the build guide

What makes Close easy to automate

Close’s data model is built for inside sales. A lead is the company, and it contains contacts, opportunities, tasks and a timeline of activities: calls, emails, SMS, meetings and notes. Because calls and emails happen inside Close, the conversation data an AI workflow needs is already in the CRM rather than scattered across a dialer and an inbox.

The REST API is consistent and well documented. Authentication for internal workflows is a single API key sent as the Basic auth username. Webhook subscriptions can filter on object type, action and field-level conditions, and Close keeps an event log you can replay from if your endpoint was down. Close also offers an MCP server for connecting AI assistants directly.

What you do not get is a built-in n8n node, so n8n workflows use HTTP Request. Zapier has a Close app, which covers simple triggers and actions.

The AI layer

Close teams make many calls, which means the biggest win is turning conversations into structured data and better follow-ups:

  • Call capture. Summaries, objections, next steps and opportunity changes written back after each call. See how to automate CRM data entry from sales calls.
  • Coaching. Consistent, private feedback on calls against your own playbook. Our AI sales call coaching guide covers how to keep it useful rather than punitive.
  • Drafting. Follow-ups written from the actual conversation.
  • Prioritization. Qualification and enrichment so reps call the right leads first.

The model writes notes and custom fields. Opportunity status and value changes are proposed, not applied, until the team trusts the workflow.

Limits to plan around

  • Rate limits per endpoint group. Limits differ by endpoint group, apply per API key, and are capped again at the organization level. Respect the reset value in 429 responses.
  • Event ordering. Close consolidates events and does not guarantee order. Fetch the current record state rather than trusting the order webhooks arrive in.
  • Webhook backlog. If your endpoint fails for a long time, events back up and Close will eventually pause the subscription. Monitor delivery.
  • Custom fields. Custom fields are referenced by IDs in the API. Map them once and keep the mapping in the workflow.
  • Plan features. Calling, recording and some automation features vary by plan. Check whether your plan includes API access and recordings before scoping call workflows.

How Lumeor typically connects it

We create one API key per workflow on a dedicated integration user, subscribe to call and lead events with filtered webhooks, and run the logic in n8n with HTTP Request nodes. Transcription and summarization run through Claude or OpenAI’s GPT models with fixed output fields. When a team calls through a separate dialer, we connect that system too, such as Aircall, and match calls to Close leads by phone number.

Frequently asked questions

Does n8n have a Close node?

No. n8n does not ship a built-in Close node, so we connect through the HTTP Request node with the API key as a Basic auth credential. Close's API is clean and consistent, so this adds little effort. Zapier does have a Close app for simpler triggers and actions.

How do I get Close call recordings into an AI workflow?

Call activities in the API include recording information when recording is enabled for your team. A workflow listens for new call activity through a webhook, fetches the recording or transcript, sends it to a model, and writes the summary back as a note. Confirm your recording settings and consent rules for the states you call.

Does Close already have AI features?

Close has been adding its own AI features, such as call summaries, and they are worth trying first if your plan includes them. A custom workflow is worth it when you want your own scoring rubric, outputs written to specific custom fields, or data combined from systems outside Close.

Do I need a separate API key per integration?

It is a good idea. Each key has its own rate limit and the organization limit is higher than any single key, so separate keys stop one heavy job from slowing the others. It also makes it easy to revoke a single integration without breaking the rest.

Next steps

01 Browse free n8n templates Fifty starter workflows you can import today. No email required. 02 Take the 5-minute AI readiness assessment Get a score and a recommended first project in your browser. 03 Get a written scope and estimate Tell us what you want to automate. We reply within two business days.

Keep reading