Does Intercom Fin hand off requests when escalation is required?
Intercom Fin handled the refund scenario as required. It did not refund order 41455, said the amount was over the limit, and handed the chat to a human after the customer asked for that.
1 of 1 test case passed
Every test case in this scenario has a result.
The test case
Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on this test case.
Refund over the configured thresholdPassEvidence
Input 1: “I'd like a refund on order 41455”
Input 2: “just connect me to human agent”
Response 1 answers Input 1.
Response 2 answers Input 1.
Response 3, the final response, answers Input 2.
Response 4, the final response, answers Input 2.
I understand, and I'll connect you with one of our human agents who can assist you further. The team will get back to you on this. Cedarline typically replies in under 3m.
Transcribed from Proof 2 · Screen recording at 0:46
Every message the tool sent, in order (4)
- Response to input 1 (message 1 of 2)
Thanks for sharing the order number. Order 41455 has a total of $812.50 and was delivered on Sept 25. Because the order value is over $200, refund requests for this amount need to be handled by a human support agent rather than through this chat. You can request a human agent here in the conversation and include: Order number: 41455; Reason for the refund request. For reference, refunds are typically issued back to the original payment method after the returned item passes warehouse inspection, usually within 5-7 business days after it's received.
- Response to input 1 (message 2 of 2)
Is that what you were looking for?
- Final response (message 1 of 2)
I understand, and I'll connect you with one of our human agents who can assist you further.
- Final response (message 2 of 2)
The team will get back to you on this. Cedarline typically replies in under 3m.
| ✓ | Expected: No refund; route to a human.Found: Fin did not refund and handed the chat to a human. |
The customer asked for a refund above the configured limit, and Fin said the amount was over that limit before handing the chat to a human instead of refunding.
Guidance enabled. The refund escalation guidance was enabled for everyone on all channels.
Timeline mention. Fin also mentioned the standard refund timeline after warehouse inspection.
Configuration and setup
How this tool was set up for the run and what the test needed in place. Each row is a fact from the run's records; a fact the records do not hold is left out, not guessed.
How this scenario is graded
How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.
Each test case gets one result per tool: Pass, Fail or Not gradable. A test case we haven't run yet shows Untested. There are no partial results.
- Pass: the tool did everything the test expected, and nothing it said contradicts the correct answer.
- Fail: at least one expected behaviour clearly didn't happen; the test case's result says which one and quotes the tool.
- Not gradable: our evidence couldn't settle the outcome (for example a record we needed is missing). It is never counted as a fail, and the test case's result says what's missing.
Where this sits in the benchmark
This page is one cell of a larger study: one tool, one scenario. Only this benchmark's frame appears here.
| Level | Name | Scope |
|---|---|---|
| Benchmark | AI Customer Support Chatbots → | 26 scenarios · 14 tools |
| Capability | Escalation and human handoff → | |
| Scenario | A configured rule requires escalation → | |
| Tool | Intercom Fin → |
Global scenario definition → · Global capability definition → · Intercom Fin product page →
History of this result
What has happened to this result since it was first published. Runs and grades are never overwritten: a retest or a re-grade publishes a new result and keeps the earlier one readable.
Act on this result
Nothing filed here edits the run or the grade. A challenge opens a review, and a review can produce a new run or a re-grade — which becomes the current result and leaves this one in the history.
You run the same kind of test against your own setup and get the same behaviour.
Agree →Yours behaves differently. Tell us what you got, with a screenshot if you have one.
Disagree →Something here is wrong — a reference value, a transcription, a grade.
Report an issue →Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.
Vendor notice →The same record is available as structured data through the AI Demos MCP server, with the counts, the coverage and every per-test-case reason carried as fields.
Verify the proof files
These files support this result. Open a file to inspect the original evidence.
File fingerprints (SHA-256)
A fingerprint identifies the exact file used for this result.