Can camelAI find records when a question uses a synonym for a stored value?
camelAI found the orders despite the wording mismatch: asked for orders paid with a gift voucher, it searched the payment method for both gift card and gift voucher and returned the 14 gift card orders, two of them cancelled, as its answer. No empty result was presented.
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.
Category asked by the user's synonymPassEvidence
“Show me all orders paid with a gift voucher.”
Answer: 14 gift card orders
I found 14 orders recorded with payment method gift_card in customer_support_db. Two are cancelled, so they may not represent completed payments.
Note under the table
The payment field is gift_card ; the source doesn't have a separate gift_voucher value in these results.
Transcribed from Proof 1 · Screen recording at 0:16
Transcribed from Proof 1 · Screen recording at 0:22
Every message the tool sent, in order (3)
- Query behind the answer · not part of the answer
SELECT o.order_number, o.placed_at, c.name AS customer, o.status::text AS status, o.item_summary, o.total_amount, o.currency, o.payment_method FROM orders o LEFT JOIN customers c ON c.id=o.customer_id WHERE lower(o.payment_method) IN ('gift_card','gift_voucher') ORDER BY o.placed_at DESC, o.order_number - Answer: 14 gift card orders
The final response, printed above.
- Note under the table
The final response, printed above.
| ✓ | Expected: It returns the matching orders; an empty result presented as the answer is the failure, since it reads to the user as a fact.Found: It returned the 14 orders stored with a gift card payment method and presented them as the answer, the same 14 orders, total and two cancellations the reference query returns. |
An empty result presented as the answer would be the failure. A read-only query on the same database returns 14 gift card orders totalling 16,806.08, two of them cancelled; camelAI's answer lists the same 14 orders and flags the same two cancellations, so this is a Pass.
Query matched both values. The generated query matched both the gift card value and a literal gift voucher value; no order is stored as gift voucher.
Stated its interpretation. Under the table, camelAI said the payment field holds gift card and the source has no separate gift voucher value, so the mapping is visible to the user.
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 Database Agents → | 28 scenarios · 11 tools |
| Capability | Question Answering → | |
| Scenario | The user's wording doesn't match how the values are stored → | |
| Tool | camelAI → |
Global scenario definition → · Global capability definition → · camelAI 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.