Integration Tutorial April 13, 2026

Add Customer History Memory to Intercom Bots

Intercom bots answer the current question but forget the last one. REM Labs gives them persistent memory -- syncing conversation history, contact data, and help articles so your bot knows who it is talking to, what they asked before, and what your docs say about it.

What Gets Synced

The Intercom integration pulls conversation and contact data to build a comprehensive customer memory:

Conversations -- full message threads including customer messages, agent replies, and bot responses
Contacts -- user and lead profiles with custom attributes, tags, and segments
Help Center articles -- all published help articles and collections for knowledge base search
Notes -- internal team notes attached to contacts
Tags and segments -- stored as filterable metadata for scoped retrieval

Connect via OAuth

In the REM Labs Console, go to Integrations and click Connect Intercom. Authorize via Intercom OAuth. The initial sync imports your conversation history, contact profiles, and help articles.

Import via API

curl -X POST https://api.remlabs.ai/v1/memory/sync/intercom/import \
  -H "Authorization: Bearer sk-slop-..." \
  -H "Content-Type: application/json" \
  -d '{
  "intercom_token": "dG9rOm...",
  "include_conversations": true,
  "include_contacts": true,
  "include_articles": true,
  "conversation_state": "closed",
  "sync_mode": "incremental"
}'

Setting conversation_state: "closed" imports only resolved conversations, which is useful for building a knowledge base from successful resolutions. Use "all" to include open conversations too.

Real-Time Bot Memory

The real power is using REM Labs as a memory backend for your Intercom bot. When a customer opens a new conversation, query REM Labs for their history before generating a response.

curl -X POST https://api.remlabs.ai/v1/integrations/intercom/search \
  -H "Authorization: Bearer sk-slop-..." \
  -H "Content-Type: application/json" \
  -d '{
  "query": "billing issue with annual subscription",
  "contact_id": "usr_63a1b2c3d4",
  "limit": 5
}'

This returns the customer's previous conversations about billing, the help article about subscription management, and the internal note the support team left about their account status -- all ranked by relevance to the current issue.

Help Center as Knowledge Base

Your Intercom Help Center articles are indexed with semantic search. When a customer asks a question, REM Labs finds the relevant article even if the customer uses completely different terminology than the article title. "How do I change my payment info" matches the article titled "Updating Billing Details in Your Account."

API Endpoints Reference

Endpoint	Method	Description
`/v1/memory/sync/intercom/import`	POST	Import conversations, contacts, and articles
`/v1/integrations/intercom/webhook`	POST	Configure real-time conversation sync
`/v1/integrations/intercom/status`	GET	Check sync status and conversation count
`/v1/integrations/intercom/search`	POST	Search within the Intercom namespace

Contact-scoped retrieval: Pass a contact_id to scope search results to a specific customer's history. This is the key to personalized bot responses -- the bot retrieves only that customer's past conversations and account context. See Intercom integration docs.

Give your Intercom bot a memory

Free tier. Persistent customer context across every conversation.

Get started free →