CHATBOTS

Intercom Self-Serve Bot That Answers from Docs Before Offering a Call

Answers onboarding questions in Intercom from your knowledge base via RAG, and only when it can't confidently resolve the question does it offer a booked Zoom activation call.

CategoryChatbots
EngineSim + Paperclip
Difficultyadvanced
Triggerevent
Steps6
Setup~25 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew Intercom onboarding question receivedIntercomIntercom
  • ActionRetrieve relevant passages from help docsConfluenceConfluence
  • ActionCompose answer and self-score confidenceOpenAI
  • LogicBranch on answer confidence
  • ActionCreate Zoom call when confidence is lowZoomZoom
  • OutputReply in Intercom with answer or call offerIntercomIntercom

What it does

This bot deflects the easy questions and reserves human time for the hard ones. Every inbound Intercom onboarding question is answered using retrieval over your help docs. The bot grades its own answer confidence — high-confidence answers ship directly, while low-confidence or repeated questions trigger an offer to book a live Zoom call.

When to use it

Use this when most onboarding questions are answerable from docs but a meaningful tail genuinely needs a human. It keeps the call calendar reserved for users who actually need it instead of every "how do I..." message.

How it works

  1. 1A new Intercom onboarding question fires the trigger.
  2. 2Relevant doc passages are retrieved from your Confluence space.
  3. 3OpenAI composes an answer and self-scores its confidence.
  4. 4A logic branch sends confident answers straight back to Intercom.
  5. 5When confidence is low, the bot apologizes for the gap and offers a call.
  6. 6On acceptance, a Zoom meeting is created and posted into the thread; the conversation is tagged for the success team.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect IntercomConversations, contacts, articles.
  2. 2
    Connect ConfluenceSpaces, pages, blueprints.
  3. 3
    Connect OpenAIModels, embeddings, files.
  4. 4
    Connect ZoomMeetings, recordings, transcripts.
  5. 5
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  6. 6
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  7. 7
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.