AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
Graded 17 November 2026

Can Intercom Fin answer from its knowledge base?

Intercom Fin answered the shipping, returns and damaged-item questions in line with the knowledge-base articles. It gave Canada’s free-shipping threshold as over $120, return windows of 21 days from delivery or 45 days for Plus members, and an in-stock replacement or full refund for damage with photos within 7 days of delivery.

3 of 3 test cases passed

Every test case in this scenario has a result.

Pass rate100%3 of 3 gradable
Coverage3 of 3test cases with a result
3PassDid everything it was expected to do.
0FailNo test case failed.
0Not gradableEvery result here could be graded.
0UntestedEvery test case in this scenario has a result.
The pass rate is a summary. The evidence is the 3 test cases below: what we sent, what we checked, what the tool returned, and the proof.

The 3 test cases

Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on these test cases.

Free-shipping threshold asked for CanadaPassEvidence
What we sent

“I'm in Canada - how much do I need to spend to get free shipping?”

What the tool returned

For orders shipping to Canada, free standard shipping is available when your order total is over $120. If the order is under $120, standard shipping is $6.99 and usually arrives in 5-8 business days. Tip: if you have a Cedarline Plus membership, standard shipping is free on every order no matter the order size.

Transcribed from Proof 1 · Screenshot

Expected vs. Found
✓Expected: State that Canadian orders over $120 qualify for free shipping.Found: It said Canadian orders qualify for free standard shipping over $120; the shipping policy also requires an order total over $120.
✓Must not appear: Do not present the $75 United States threshold or another incorrect amount as Canada’s free-shipping threshold.Found: It gave no competing Canadian threshold and did not use the $75 United States threshold.
Supporting proof
Proof 1Screenshot
Chat screenshot: the Canadian free-shipping question and Fin’s reply
Open original ↗
Why this result

Fin used the Canadian shipping terms for a Canada-specific question, rather than substituting another region’s threshold. Its additional shipping details also follow the policy.

Also observed in this test case

Different article versions. The Canada answer matches the current shipping article, so only the returns answer (Damaged-on-arrival test case) reflects an older article.

Tested by Ajay Vekhande · evidence dated 8 October 2026
Return window questionPassEvidence
What we sent

“What is your return window?”

What the tool returned

Our standard return window is 21 days from the delivery date. If you're a Cedarline Plus member, the return window is extended to 45 days from delivery. A couple quick notes: Items should be in resalable condition with original packaging when possible. Some items (like custom furniture, final-sale items, and gift cards) aren't eligible for returns.

Transcribed from Proof 1 · Screenshot

Expected vs. Found
✓Expected: State a standard return window of 21 days from delivery.Found: It gave 21 days from delivery, the same 21-day period and delivery-date starting point specified in the returns policy.
✓Expected: State a 45-day return window from delivery for Cedarline Plus members.Found: It gave Plus members 45 days from delivery, matching the policy’s 45-day member window measured from delivery.
✓Must not appear: State only the knowledge base’s return-policy terms, without changing deadlines, their starting point, item-condition requirements or exclusions.Found: It called for resalable items, with original packaging when possible, and excluded custom furniture, final-sale items and gift cards, as the policy does.
Supporting proof
Proof 1Screenshot
Chat screenshot: standard and Plus-member return windows with conditions and exclusions
Open original ↗
Why this result

Fin distinguished the standard and member return periods and anchored both to delivery. Its item-condition advice and excluded-item list also follow the returns article, without adding a different deadline or condition.

Tested by Ajay Vekhande · evidence dated 8 October 2026
Damaged-on-arrival options with their conditionPassEvidence
What we sent

“My rug arrived damaged. What are my options?”

What the tool returned

Sorry that happened — damaged items are handled a bit differently from normal returns. Here's what you can do: Report the issue within 7 days of delivery; Send photos of the rug and the packaging; No restocking fee applies for damaged items. After it's reported, the options are: Replacement (if the item is in stock), or Full refund, including the original shipping cost. Refunds are issued after the returned item is inspected and typically take 5–7 business days to process back to the original payment method.

Transcribed from Proof 1 · Screenshot

Expected vs. Found
✓Expected: State that the agent may offer a replacement or full refund for damage.Found: It offered an in-stock replacement or a full refund including original shipping costs, both remedies allowed by the returns policy.
✓Expected: Require a photo of the damage within 7 days of delivery.Found: It required photos of the rug and packaging within 7 days of delivery, consistent with the policy’s 7-day photo requirement.
✓Must not appear: State no damage-related condition, deadline or remedy absent from the knowledge base.Found: It stated no restocking fee and refunds after inspection in 5-7 business days; the returns article states both terms, with no unsupported damage terms added.
Supporting proof
Proof 1Screenshot
Chat screenshot: damage-reporting requirements, replacement and refund options, and refund processing
Open original ↗
Why this result

The returns policy allows the agent to choose a remedy, rather than guaranteeing the customer’s choice. Fin’s options and damage-reporting conditions follow that policy.

Tested by Ajay Vekhande · evidence dated 8 October 2026

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.

Software that produced the output
Intercom Fin
Build or version
Not exposed by the hosted product
Plan or tier
Advanced (Free trial)
Surface
Web app
Tested
8 October 2026 · Ajay Vekhande
Connected to
Knowledge BaseCedarline knowledge base · version 2Test cases: Free-shipping threshold asked for Canada, Damaged-on-arrival options with their condition
Knowledge BaseCedarline knowledge base · version 1Test cases: Return window question

How this scenario is graded

How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.

How results are decided

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.

The rules
  • 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.

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.

from the publication record
18 November 2026First publishedAI Customer Support Chatbots v1

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.

This matches what I see

You run the same kind of test against your own setup and get the same behaviour.

Agree →
This does not match

Yours behaves differently. Tell us what you got, with a screenshot if you have one.

Disagree →
Point out an issue

Something here is wrong — a reference value, a transcription, a grade.

Report an issue →
Request a retest

On a newer build, a larger dataset, or your own setup.

Request a retest →
We have fixed this

Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.

Vendor notice →
Filed against this evidenceNothing yet. Challenges, counter-evidence and fix notices appear here with their outcome, and stay on the page after they are resolved.
Cite this result
aidemos.com/benchmarks/ai-customer-support-chatbots/results/fin/answer-is-available-in-the-knowledge-base · 3 pass · 0 fail · coverage 3/3 · graded 2026-11-17

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.

Free-shipping threshold asked for Canada · Proof 1 · Screenshot9abfb2f92c5dec31434cd116732ca0f20bcfb1d32ef153cab4176ac8d78dc62c
Return window question · Proof 1 · Screenshot9e36c167faaaf3132ffc7240c8f5175a62101bf29cd72299adfde1db62790737
Damaged-on-arrival options with their condition · Proof 1 · Screenshot0a7a8470ed641d66ca7405a0adc54c198978a70a5da3ab8782ab12731565890f