AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
Graded 23 October 2026

Can FutureSmart Document Intelligence use changed existing content?

FutureSmart Document Intelligence passed this changed-content scenario. Before the refresh it answered 11 working days; after the old B02 was deleted and the replacement ingested, it answered 6 working days, with the new file first and the old value absent.

1 of 1 test case passed

Every test case in this scenario has a result.

Pass rate100%1 of 1 gradable
Coverage1 of 1test case with a result
1PassDid 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 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 counts this row.

Indexed value changed at the source, then queried after a refreshPassEvidence
What we sent

We asked for the Bergen shipping lead time, deleted the old B02 and re-added the replacement under the same filename, then asked the same lead time again.

What the tool returned

The shipping lead time to the Bergen depot is **6 working days** via road and ferry groupage, starting from courier collection. Shipments handed over after the 15:30 cut-off travel the following working day.

Transcribed from Proof 1 · Screenshot

Every message the tool sent, in order (2)
  1. Response to input 1

    The shipping lead time from the central store in Bristol to the Bergen depot is **11 working days** via road and ferry groupage. This assumes standard groupage service; consignments handed over after the 15:30 cut-off travel the following working day. Dangerous goods containing a lithium cell above the threshold require an additional **2 working days**.

  2. Final response

    The shipping lead time to the Bergen depot is **6 working days** via road and ferry groupage, starting from courier collection. Shipments handed over after the 15:30 cut-off travel the following working day.

Output file (text)
Text evidence file: before-and-after Bergen lead-time answers and checks
Text654 lines
TC99 (Scenario S89) — evidence file
ClickUp: https://app.clickup.com/t/9014651757/86bbtm6w6
Capability: C36 Knowledge Sync — mutation slice, change step (MS-v1 -> MS-v2)

Verbatim query: What is the shipping lead time to the Bergen depot?
Reference answer: 6 working days comes back; 11 working days must NOT come back

Same mutation-slice collection as TC98 (never mixed with the SS-v1 stable slice).
Change applied via delete + re-add under the IDENTICAL filename (the API has no
PATCH/update-in-place endpoint for a document -- PATCH .../documents/{id} is 405,
only DELETE is allowed -- so this is the correct way to operationalize an in-place
content 'change' without silently re-testing 'add' via a rename). Old B02 document_id
deleted, new B02_shipping_lead_times.pdf (same filename, Bergen row 11 -> 6 working
days) ingested and confirmed indexed. Per ground truth: MS-v2 digest a1de5a13d3f0d9a1
(5 docs). The old value (11 working days) appears nowhere else in the corpus, so its
continued return could only mean the index hadn't caught up.

==============================================================================
BEFORE — queried at MS-v1 (old B02, Bergen = 11 working days)
==============================================================================
status: 200
answer: The shipping lead time from the central store in Bristol to the Bergen depot is **11 working
days** via road and ferry groupage. This assumes standard groupage service; consignments handed over
after the 15:30 cut-off travel the following working day. Dangerous goods containing a lithium cell
above the threshold require an additional **2 working days**.
contains '11 working days' in any retrieved section: True

all 6 retrieved sections (full content), sorted by score:

score=1.0000  document_id=0b4ff443-a8ea-4904-a8b3-d93f468ff435
File Name: B02_shipping_lead_times.pdf

Extracted Text:

Fenwake Instruments Ltd

# Shipping Lead Times

Central store, Bristol - outbound to depots and customer sites

## How to read this

Lead times are quoted from the courier collection, not from the moment a stores request is raised.
Anything handed over after the 15:30 cut-off travels the following working day, and the lead time
starts then.

<table>
  <thead>
    <tr>
      <th>Destination</th>
      <th>Route</th>
      <th>Lead time</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Aberdeen</td>
<td>Road, own vehicle</td>
<td>3 working days</td>
    </tr>
<tr>
      <td>Lerwick</td>
<td>Road and ferry</td>
<td>4 working days</td>
    </tr>
<tr>
      <td>Gothenburg</td>
<td>Road and ferry, groupage</td>
<td>7 working days</td>
    </tr>
<tr>
      <td>Esbjerg</td>
<td>Road and ferry, groupage</td>
<td>9 working days</td>
    </tr>
<tr>
      <td>Bergen</td>
<td>Road and ferry, groupage</td>
<td>11 working days</td>
    </tr>
  </tbody>
</table>

Lead times to the Nordic depots assume a groupage service. A dedicated vehicle is quoted on request
and is normally quicker, but is not held against these figures.

## Dangerous goods

Any consignment containing a lithium cell above the threshold needs a signed packing declaration and
travels on a different service. Add two working days to the figures above, and involve a trained
shipper before booking. Two people in the building hold the certification.

## Returns

Inbound lead times are not the outbound figures in reverse. A return from a depot travels on the
next scheduled groupage movement rather than on collection, so plan a return around the movement
rather than around the courier.

## Who maintains this

The stores controller reviews these figures against the carrier's performance each quarter and
changes them here when the carrier's own schedule changes.

Central store, Bristol

Page 1

score=0.6667  document_id=e747664e-aa91-4ec3-82fd-9365709af4e4
File Name: B05_tromso_depot_bulletin.md

Extracted Text: # Depot bulletin - a new depot in the region

For the service desk, the Bergen depot and central stores.

## What is opening

The Tromso depot opens on 14 October 2026 and will hold four Tolvan C5 loan units and a full Brannoc
spares kit. It is the third depot in the network and the first
above the Arctic Circle.

## Why

Response to the northern fields has been limited by the sailing time from
Bergen rather than by engineering. Holding loan stock and a spares kit locally
removes the leg that has cost the most time on recent tickets.

## Staffing

One engineer and a shared stores controller, working 07:00 to 15:00 local time,
Monday to Friday. Out of hours the northern sites continue to page the existing
duty engineer rota; the new depot does not change how a fault is reported.

## What to do differently

Stop routing northern loan requests through Bergen once the depot opens. Ring
the new depot directly. Bergen keeps its own loan pool and its own spares kit,
and nothing moves out of Bergen to stock the new site.

## What is not changing

Calibration stays in the laboratory. The new depot does not calibrate, does not
break a calibration seal, and does not open a pressure housing.

score=0.4500  document_id=d9467dab-9f21-4868-8ddb-b6ca407f7634
File Name: B03_approved_supplier_list.docx

Extracted Text: # Approved Supplier List

*Fenwake Instruments Ltd - *parts fitted to instruments in service*

## 1. Scope

This list names the suppliers approved for parts fitted to instruments in service. A part from a
supplier not on this list is not fitted, whatever the lead time pressure, because the part becomes
part of a calibrated instrument.

## 2. Single-source parts

The sole approved supplier of the H-clamp bracket is Ardwyn Fabrication (Newport), quoted per batch
of fifty. The bracket is machined to our drawing and no second source has been qualified, so a
shortage here stops the fitting rather than being worked around.

Where a single-source part is short, the stores controller tells the service desk lead the same day.
A substitution is authorised by the head of service operations in writing, and the instrument record
carries the substitution.

## 3. Approved suppliers

<table>
  <tr>
    <th>
      <b>Part</b>
    </th>
    <th>
      <b>Supplier</b>
    </th>
    <th>
      <b>Location</b>
    </th>
  </tr>
<tr>
    <td>
      H-clamp bracket
    </td>
<td>
      Ardwyn Fabrication
    </td>
<td>
      Newport
    </td>
  </tr>
<tr>
    <td>
      Sacrificial anode set
    </td>
<td>
      Trellech Metals
    </td>
<td>
      Chepstow
    </td>
  </tr>
<tr>
    <td>
      Anti-fouling guard
    </td>
<td>
      Coldbrook Polymers
    </td>
<td>
      Taunton
    </td>
  </tr>
<tr>
    <td>
      Cable gland kit
    </td>
<td>
      Sedgemoor Sealing
    </td>
<td>
      Bridgwater
    </td>
  </tr>
<tr>
    <td>
      Armoured deck cable
    </td>
<td>
      Nailsea Cable Works
    </td>
<td>
      Nailsea
    </td>
  </tr>
</table>

## 4. Qualifying a new supplier

score=0.3929  document_id=de5dbe95-5830-434b-92bf-9a0229d26408
File Name: B01_oncall_rota_notes.md

Extracted Text: # On-call rota - working notes

For the service desk lead. Not a contractual document.

## Shape of the rota

The duty engineer rota runs Monday to Monday, handing over at 09:00. Four
engineers are on the rota at any time, which puts each of them on call one week
in four. An engineer who is on call does not travel offshore in the same week.

## Cover for absence

Where an engineer on the rota is absent, the service desk lead swaps rather
than doubles: the week is given to the next engineer in the cycle and the
absent engineer takes theirs. Doubling a week has caused more missed pages than
any other cause and is not done.

On a statutory non-working day the rota does not change. The standby allowance
for that week is paid at the enhanced rate set out in the handbook.

## Handover

Handover is a call, not a message. The outgoing engineer walks the incoming one
through every ticket still open, what has been tried, and what the customer has
been told. A ticket that changes hands without that call is the single most
common source of a customer being told two different things.

## Paging

The out-of-hours number pages the engineer on call. Where the page is not
acknowledged, it escalates to the service desk lead automatically, and then to
the head of service operations. Test the pager at the start of every week
before the handover call ends.

score=0.3667  document_id=d9467dab-9f21-4868-8ddb-b6ca407f7634
## 4. Qualifying a new supplier

A new supplier is qualified on a first-article inspection and a documented material certificate.
Qualification is recorded against the part, not against the supplier as a whole: a supplier approved
for one part is not thereby approved for another.

## 5. Review

The stores controller reviews this list each quarter with the workshop manager. A supplier that has
missed two consecutive delivery commitments is reviewed regardless of the schedule.

score=0.3095  document_id=f10c469d-35b5-49c5-8969-fe845d247668
File Name: B04_site_access_notes.txt

Extracted Text: FENWAKE INSTRUMENTS LTD
Site access - standing notes for engineers
Central store and service desk. Not a contractual document.

PERMITS

   Most customer sites need a permit raised before the day, not on it. Raise it
   through the customer's own system and take the reference with you; a permit
   that exists only in an inbox is a permit that will not open a gate.

   A permit is site-specific and person-specific. An engineer who swaps onto a
   job at short notice needs their own, and the swap is the reason most access
   is refused at the gate.

INDUCTIONS

   Water treatment sites and harbour authorities run their own inductions,
   valid for twelve months. The service desk keeps the expiry dates against
   each engineer. An induction that expires mid-job is treated as expired for
   the whole job.

TIDE AND ACCESS WINDOWS

   Intake structures are reachable only around low water. Plan the visit around
   the window and tell the customer which window you are working to, so that a
   delayed start is a conversation rather than a surprise.

KEYS AND CODES

   Never write a gate code on a job sheet. Codes are held by the service desk
   and read out on the day. A code that has changed is reported the same day,
   because the next engineer will otherwise repeat the wasted journey.

VESSELS

   On a vessel, the master's instructions take precedence. Board with your own
   survival certification in date; the vessel will check it and will refuse
   boarding without it.

WHAT TO DO IF ACCESS IS REFUSED

   Do not argue at the gate. Ring the service desk, who will ring the
   customer's site contact. Record the refusal and the reason on the ticket
   before leaving, while the detail is still accurate.

==============================================================================
AFTER — old B02 deleted, new B02 ingested (same filename), confirmed indexed,
queried again at MS-v2 (Bergen = 6 working days)
==============================================================================
status: 200
answer: The shipping lead time to the Bergen depot is **6 working days** via road and ferry
groupage, starting from courier collection. Shipments handed over after the 15:30 cut-off travel the
following working day.
contains '11 working days' in any retrieved section: False  (must be False)
contains '6 working days' in any retrieved section: True  (must be True)

all 6 retrieved sections (full content), sorted by score:

score=1.0000  document_id=1435f1c3-1b3b-436f-b710-fc38861e5b0d
File Name: B02_shipping_lead_times.pdf

Extracted Text:

Fenwake Instruments Ltd

# Shipping Lead Times

Central store, Bristol - outbound to depots and customer sites

## How to read this

Lead times are quoted from the courier collection, not from the moment a stores request is raised.
Anything handed over after the 15:30 cut-off travels the following working day, and the lead time
starts then.

<table>
  <thead>
    <tr>
      <th>Destination</th>
      <th>Route</th>
      <th>Lead time</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Aberdeen</td>
<td>Road, own vehicle</td>
<td>3 working days</td>
    </tr>
<tr>
      <td>Lerwick</td>
<td>Road and ferry</td>
<td>4 working days</td>
    </tr>
<tr>
      <td>Gothenburg</td>
<td>Road and ferry, groupage</td>
<td>7 working days</td>
    </tr>
<tr>
      <td>Esbjerg</td>
<td>Road and ferry, groupage</td>
<td>9 working days</td>
    </tr>
<tr>
      <td>Bergen</td>
<td>Road and ferry, groupage</td>
<td>6 working days</td>
    </tr>
  </tbody>
</table>

Lead times to the Nordic depots assume a groupage service. A dedicated vehicle is quoted on request
and is normally quicker, but is not held against these figures.

## Dangerous goods

Any consignment containing a lithium cell above the threshold needs a signed packing declaration and
travels on a different service. Add two working days to the figures above, and involve a trained
shipper before booking. Two people in the building hold the certification.

## Returns

Inbound lead times are not the outbound figures in reverse. A return from a depot travels on the
next scheduled groupage movement rather than on collection, so plan a return around the movement
rather than around the courier.

## Who maintains this

The stores controller reviews these figures against the carrier's performance each quarter and
changes them here when the carrier's own schedule changes.

Central store, Bristol
Page 1

score=0.6667  document_id=e747664e-aa91-4ec3-82fd-9365709af4e4
File Name: B05_tromso_depot_bulletin.md

Extracted Text: # Depot bulletin - a new depot in the region

For the service desk, the Bergen depot and central stores.

## What is opening

The Tromso depot opens on 14 October 2026 and will hold four Tolvan C5 loan units and a full Brannoc
spares kit. It is the third depot in the network and the first
above the Arctic Circle.

## Why

Response to the northern fields has been limited by the sailing time from
Bergen rather than by engineering. Holding loan stock and a spares kit locally
removes the leg that has cost the most time on recent tickets.

## Staffing

One engineer and a shared stores controller, working 07:00 to 15:00 local time,
Monday to Friday. Out of hours the northern sites continue to page the existing
duty engineer rota; the new depot does not change how a fault is reported.

## What to do differently

Stop routing northern loan requests through Bergen once the depot opens. Ring
the new depot directly. Bergen keeps its own loan pool and its own spares kit,
and nothing moves out of Bergen to stock the new site.

## What is not changing

Calibration stays in the laboratory. The new depot does not calibrate, does not
break a calibration seal, and does not open a pressure housing.

score=0.4500  document_id=d9467dab-9f21-4868-8ddb-b6ca407f7634
File Name: B03_approved_supplier_list.docx

Extracted Text: # Approved Supplier List

*Fenwake Instruments Ltd - *parts fitted to instruments in service*

## 1. Scope

This list names the suppliers approved for parts fitted to instruments in service. A part from a
supplier not on this list is not fitted, whatever the lead time pressure, because the part becomes
part of a calibrated instrument.

## 2. Single-source parts

The sole approved supplier of the H-clamp bracket is Ardwyn Fabrication (Newport), quoted per batch
of fifty. The bracket is machined to our drawing and no second source has been qualified, so a
shortage here stops the fitting rather than being worked around.

Where a single-source part is short, the stores controller tells the service desk lead the same day.
A substitution is authorised by the head of service operations in writing, and the instrument record
carries the substitution.

## 3. Approved suppliers

<table>
  <tr>
    <th>
      <b>Part</b>
    </th>
    <th>
      <b>Supplier</b>
    </th>
    <th>
      <b>Location</b>
    </th>
  </tr>
<tr>
    <td>
      H-clamp bracket
    </td>
<td>
      Ardwyn Fabrication
    </td>
<td>
      Newport
    </td>
  </tr>
<tr>
    <td>
      Sacrificial anode set
    </td>
<td>
      Trellech Metals
    </td>
<td>
      Chepstow
    </td>
  </tr>
<tr>
    <td>
      Anti-fouling guard
    </td>
<td>
      Coldbrook Polymers
    </td>
<td>
      Taunton
    </td>
  </tr>
<tr>
    <td>
      Cable gland kit
    </td>
<td>
      Sedgemoor Sealing
    </td>
<td>
      Bridgwater
    </td>
  </tr>
<tr>
    <td>
      Armoured deck cable
    </td>
<td>
      Nailsea Cable Works
    </td>
<td>
      Nailsea
    </td>
  </tr>
</table>

## 4. Qualifying a new supplier

score=0.3929  document_id=de5dbe95-5830-434b-92bf-9a0229d26408
File Name: B01_oncall_rota_notes.md

Extracted Text: # On-call rota - working notes

For the service desk lead. Not a contractual document.

## Shape of the rota

The duty engineer rota runs Monday to Monday, handing over at 09:00. Four
engineers are on the rota at any time, which puts each of them on call one week
in four. An engineer who is on call does not travel offshore in the same week.

## Cover for absence

Where an engineer on the rota is absent, the service desk lead swaps rather
than doubles: the week is given to the next engineer in the cycle and the
absent engineer takes theirs. Doubling a week has caused more missed pages than
any other cause and is not done.

On a statutory non-working day the rota does not change. The standby allowance
for that week is paid at the enhanced rate set out in the handbook.

## Handover

Handover is a call, not a message. The outgoing engineer walks the incoming one
through every ticket still open, what has been tried, and what the customer has
been told. A ticket that changes hands without that call is the single most
common source of a customer being told two different things.

## Paging

The out-of-hours number pages the engineer on call. Where the page is not
acknowledged, it escalates to the service desk lead automatically, and then to
the head of service operations. Test the pager at the start of every week
before the handover call ends.

score=0.3667  document_id=d9467dab-9f21-4868-8ddb-b6ca407f7634
## 4. Qualifying a new supplier

A new supplier is qualified on a first-article inspection and a documented material certificate.
Qualification is recorded against the part, not against the supplier as a whole: a supplier approved
for one part is not thereby approved for another.

## 5. Review

The stores controller reviews this list each quarter with the workshop manager. A supplier that has
missed two consecutive delivery commitments is reviewed regardless of the schedule.

score=0.3095  document_id=f10c469d-35b5-49c5-8969-fe845d247668
File Name: B04_site_access_notes.txt

Extracted Text: FENWAKE INSTRUMENTS LTD
Site access - standing notes for engineers
Central store and service desk. Not a contractual document.

PERMITS

   Most customer sites need a permit raised before the day, not on it. Raise it
   through the customer's own system and take the reference with you; a permit
   that exists only in an inbox is a permit that will not open a gate.

   A permit is site-specific and person-specific. An engineer who swaps onto a
   job at short notice needs their own, and the swap is the reason most access
   is refused at the gate.

INDUCTIONS

   Water treatment sites and harbour authorities run their own inductions,
   valid for twelve months. The service desk keeps the expiry dates against
   each engineer. An induction that expires mid-job is treated as expired for
   the whole job.

TIDE AND ACCESS WINDOWS

   Intake structures are reachable only around low water. Plan the visit around
   the window and tell the customer which window you are working to, so that a
   delayed start is a conversation rather than a surprise.

KEYS AND CODES

   Never write a gate code on a job sheet. Codes are held by the service desk
   and read out on the day. A code that has changed is reported the same day,
   because the next engineer will otherwise repeat the wasted journey.

VESSELS

   On a vessel, the master's instructions take precedence. Board with your own
   survival certification in date; the vessel will check it and will refuse
   boarding without it.

WHAT TO DO IF ACCESS IS REFUSED

   Do not argue at the gate. Ring the service desk, who will ring the
   customer's site contact. Record the refusal and the reason on the ticket
   before leaving, while the detail is still accurate.
Open the text file ↗
Expected vs. Found
✓Expected: The new value comes back and the old value does not.Found: Before the refresh it returned 11 working days; after the refresh it returned 6 working days, and 11 working days was absent from the returned sections.
Supporting proof
Proof 1Screenshot
Notebook capture: after refresh, Bergen lead time returned as 6 working days
Open original ↗
Proof 2Screenshot
Notebook capture: before refresh, Bergen lead time returned as 11 working days
Open original ↗
Why this result

The case required the changed value to replace the old one after the refresh. The first query returned 11 working days; after the old file was removed and the replacement ingested, the same query returned 6 working days, and the old value was no longer in the returned sections.

Also observed on this row

No update-in-place. The product had no document update-in-place path; the old file was deleted and the replacement was re-added under the same filename.

Lower-ranked context. The result also retrieved lower-ranked B03 supplier-list text.

Tested by Vivek Tomar · evidence dated 8 September 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
FutureSmart Document Intelligence
Build or version
v1
Surface
API
Set up before the run
The corpus had to contain no other copy of the old value. A document already in the index was changed by replacing the old version with a new one under the same filename, then queried again.
Tested
8 September 2026 · Vivek Tomar

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 row says which 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 row 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.

LevelNameScope
BenchmarkManaged RAG →15 scenarios · 8 tools
CapabilityKnowledge Sync →
ScenarioExisting content is changed →
ToolFutureSmart Document Intelligence →

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
23 October 2026First publishedManaged RAG 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/managed-rag/results/futuresmart-document-intelligence/existing-content-is-changed · 1 pass · 0 fail · coverage 1/1 · graded 2026-10-23

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.

Output file (text) ↗text/plain; charset=utf-8 · 20 KB
Proof 1 · Screenshot ↗image/png · 436 KB
Proof 2 · Screenshot ↗image/png · 184 KB
File fingerprints (SHA-256)

A fingerprint identifies the exact file used for this result.

Output file (text)072951fbad9e95033dd11a28045811265dbb01722cf27962edd538ef2b8b695a
Proof 1 · Screenshot0a31e9f408a531cab11d275a4c80d01f884e53534d9da0dd2b04004f22a19c5b
Proof 2 · Screenshot176ac205783d2882de6ffde0d2b89524e9b3eea54123e66d5d1352a96d3646af