AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
superseded result

This is the result as it was published on 13 November 2026. A newer result replaced it on 16 November 2026; this page is kept so that anything that cited it still resolves. Read the current result →

Graded 12 November 2026

Does Dot's reported analytics match what actually happened?

Dot's History message total rose from 49 to 58 (+9) after six messages were sent, rather than the expected 55; the three extra came from sessions Dot created about missing data and counted as the user's messages. Both unanswered chats were kept with one message each, and filtering to Analysis still showed 58.

1 of 1 test case failed

Every test case in this scenario has a result.

Pass rate0%0 of 1 gradable
Coverage1 of 1test case with a result
0PassNo test case passed.
1FailMissed at least one thing it was expected to do; the reason and proof are with each test case below.
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 test case below: what we sent, what we checked, what the tool returned, and the proof.

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.

Reported totals include the unansweredFailEvidence
What we sent

A new chat was opened.

Input 1 · Revenue question: “As of September 2, 2026, what's our total revenue in the last 30 days?”

Input 2 · Revenue chat: data source reply: “customer_support_db”

Before each of Inputs 3–6: A new chat was opened.

Input 3 · Payment-method question: “What's our most common payment method?”

Input 4 · Social-media complaints question: “How many complaints have we received on social media this month?”

Input 5 · Active-subscription question: “How many customers currently have an active subscription?”

Input 6 · Net promoter score question: “What's our net promoter score for last quarter?”

What the tool returned

History message total: unfiltered Messages 58 History chat count: unfiltered 40 chats History message total: Analysis filter Messages 58 History chat count: Analysis filter 29 chats

Transcribed from Proof 1 · Screen recording at 3:58

Transcribed from Proof 1 · Screen recording at 4:03

Transcribed from Proof 1 · Screen recording at 4:08

Transcribed from Proof 1 · Screen recording at 4:13

Expected vs. Found
✗Expected: Counts must include unanswered questions and reflect six messages sent, bringing the message total from 49 to 55.Found: The total rose from 49 to 58 (+9) for six messages: 55 expected. Three extras came from sessions Dot created about missing data. Both unanswered chats showed one message each.
Supporting proof
Proof 1Screen recording
Screen recording: the question mix and History totals before and after filtering
Key moments
Open original ↗
Proof 2Captured file (PDF)
Comparison log: messages sent and History counts before and after the question mix
Open the PDF ↗
Why this result

Keeping unanswered chats was not enough: the total rose from 49 to 58 (+9) for six messages, rather than 55. Dot counted sessions it created about missing data as the user's messages; filtering to Analysis did not correct the total.

Also observed in this test case

Usable counts in the chat list. The Analysis-filtered History list contains 29 chats. For individual chats, the last-30-days revenue chat shows two messages for two sent; the payment-method chat shows one for one sent.

Tested by Anchal Rana · evidence dated 20 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
Dot
Build or version
Dot web app, build 212f51f2c (19 Oct 2026)
Plan or tier
Free
Surface
Web app
Set up before the run
The mix combined answerable revenue, payment-method and active-subscription questions with unanswerable social-media complaints and net promoter score questions. History totals were compared with a written tally of the messages sent.
History type filter
Analysis
Tested
20 October 2026 · Anchal Rana

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
13 November 2026First publishedAI Database Agents v1
16 November 2026Updated — the page was revised
The result is unchanged: 0 pass · 1 fail. The earlier result stays readable. Read the earlier result →

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-database-agents/results/dot/analytics-accuracy · 0 pass · 1 fail · coverage 1/1 · graded 2026-11-12

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.

Proof 2 · Captured file (PDF) ↗application/pdf · 70 KB
File fingerprints (SHA-256)

A fingerprint identifies the exact file used for this result.

Proof 1 · Screen recording15c8ad624937d760b8fb707df5cc3f40e84b611c2481169a94f8547ef4a62737
Proof 2 · Captured file (PDF)f857b9c83b802f8fa17a6e40df21c89cb045f26ad1c63ff98b116cb98022d343