CUSTOMER SUPPORT

Refund Edge-Case Advisor for Intercom Conversations

When a customer asks for a refund in Intercom, an agent checks the request against your refund policy, drafts a compliant reply for the rep.

CategoryCustomer Support
Enginepaperclip
Difficultyintermediate
Triggerevent
Steps6
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew Intercom refund conversationIntercomIntercom
  • ActionFetch charges and refund history from StripeStripeStripe
  • ActionEvaluate request against refund policyOpenAI
  • LogicBranch: in-policy vs. supervisor edge case
  • ActionPost drafted reply as Intercom noteIntercomIntercom
  • OutputFlag edge cases to Slack supervisor channelSlack

What it does

It turns every inbound refund request into a ready-to-send draft plus a clear verdict: in-policy, deny, or needs a supervisor. The agent reads the conversation, pulls the purchase context from Stripe, and applies your written refund policy so reps stop guessing on tricky cases.

When to use it

Use it when refund questions land in Intercom and your team needs consistent, policy-aligned answers — especially for edge cases like out-of-window requests, partial refunds, or repeat refunders that a junior rep shouldn't decide alone.

How it works

  1. 1A new Intercom conversation tagged or classified as a refund request triggers the flow.
  2. 2The agent pulls the customer's recent charges and refund history from Stripe for context.
  3. 3An OpenAI reasoning step compares the request to your refund policy and assigns a verdict with a rationale.
  4. 4A branch checks whether the case is clearly in-policy or an edge case (exceptions, high amount, prior refunds).
  5. 5In-policy cases get a drafted reply posted as an Intercom note for one-click send.
  6. 6Edge cases are posted to a Slack supervisor channel with the draft and the policy reasoning for human approval.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect IntercomConversations, contacts, articles.
  2. 2
    Connect StripeCustomers, subscriptions, payments.
  3. 3
    Connect OpenAIModels, embeddings, files.
  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.