CHATBOTS

Voice-reply bot with Slack escalation summary for hard questions

Voices answers to accessibility-mode Intercom visitors, but when the LLM is unsure it posts a spoken holding message and sends a Slack summary so a human can take over fast.

CategoryChatbots
Enginesim
Difficultyadvanced
Triggerwebhook
Steps6
Setup~25 min

How it runs

The automated pipeline, trigger to output.

  • TriggerIntercom accessibility-mode message webhookIntercomIntercom
  • ActionAnswer with OpenAI + confidence scoreOpenAI
  • LogicBranch on confidence: answer vs escalate
  • ActionVoice answer or holding message with ElevenLabsElevenLabsElevenLabs
  • ActionPost audio reply to IntercomIntercomIntercom
  • OutputOn escalation, send summary + link to SlackSlack

What it does

This is a voice bot that knows its limits. It answers confident questions with an ElevenLabs clip, and when a question is ambiguous, sensitive, or low-confidence it plays a short spoken holding message to the visitor while alerting a support channel in Slack with a tidy summary and a deep link to the conversation.

When to use it

Use it when accessibility-mode visitors must never hit a dead end. It keeps the voice experience for easy questions while guaranteeing a human picks up the hard ones quickly, with full context already summarized.

How it works

  1. 1An Intercom webhook fires on a new accessibility-mode message.
  2. 2OpenAI answers and also returns a confidence score and an escalation flag.
  3. 3A logic step branches on confidence. High confidence continues to a full voice answer; low confidence takes the escalation path.
  4. 4On the confident path, ElevenLabs voices the answer and the output step posts it to Intercom.
  5. 5On the escalation path, ElevenLabs voices a brief holding message, and the output step posts it to Intercom while sending Slack a summary of the question with a link for a teammate to take over.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect IntercomConversations, contacts, articles.
  2. 2
    Connect OpenAIModels, embeddings, files.
  3. 3
    Connect ElevenLabsText-to-speech, voice cloning.
  4. 4
    Connect SlackChannels, DMs, threads, mentions.
  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.