AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
Graded 20 November 2026

Can Vectorize.io read a cleanly scanned document accurately?

Vectorize.io turned a one-page scanned calibration certificate, whose text existed only as an image, into readable text and returned it as the top passage for a question about it. That passage gives the as-found bias as -0.037 psu, the figure on the certificate.

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 is based on this test case.

Fact that exists only as pixelsPassEvidence
What we sent

“What as-found bias is recorded on calibration certificate C-77413?”

What the tool returned
Textthe whole file · cited lines 1–27 and 29–201 marked
Query: What as-found bias is recorded on calibration certificate C-77413?
average_relevancy=0.2001  ndcg=1.0000

--- rank 1 by relevancy ---
similarity=0.7468  relevancy=0.5140  source=06c5221a-8323-49d4-8bf0-02db17c0c558/4dece290-11ac-46c2-8d2e-99a00b6e2193/A03_calibration_certificate_scan.pdf
INSTRUMENT  **Tolvan A3 conductivity-temperature sensor** SERIAL NUMBER  **FW-A3-40871** CUSTOMER  **Severn Estuary Monitoring Partnership** ORDER REFERENCE  **SEMP/PO/9134** DATE OF CALIBRATION  **3 March 2026** NEXT CALIBRATION DUE  **3 January 2028** PROCEDURE  **FWL-CAL-114, issue 6** AMBIENT  **20.4 C, 61 % RH**

---

## Results

Salinity channel, checked against reference standard SR-2 at four points

| Reference point | Reference value | As found | As left |
| --------------- | --------------- | -------- | ------- |
| Point 1 | 5.000 psu | 4.961 psu | 4.997 psu |
| Point 2 | 15.000 psu | 14.964 psu | 14.996 psu |
| Point 3 | 25.000 psu | 24.962 psu | 24.995 psu |
| Point 4 | 35.000 psu | 34.965 psu | 34.997 psu |

## Summary

As-found bias, mean of four points: -0.037 psu As-left bias, mean of four points: -0.004 psu Expanded uncertainty (k = 2): 0.008 psu The instrument was adjusted after the as-found check and passed.

Conductivity cell and thermistor were cleaned before the as-found run.

Signed, calibration engineer  M. Ottaway

--- rank 2 by relevancy ---
similarity=0.6690  relevancy=0.4865  source=06c5221a-8323-49d4-8bf0-02db17c0c558/4dece290-11ac-46c2-8d2e-99a00b6e2193/A10_tolvan_a3_product_overview.md
# Tolvan A3 - product overview

The Tolvan A3 is a conductivity and temperature sensor built for
continuous deployment in estuarine and coastal water. It is the instrument most
of our customers meet first, and most of the questions the service desk answers
are about this one.

## What it measures

The instrument reports conductivity, temperature and a derived practical
salinity. Conductivity is measured with an inductive cell, which has no exposed
electrodes to foul or to wear, and temperature with a thermistor mounted in the
same flow path so that the two readings describe the same parcel of water.
Pressure is not measured; where a depth record is needed, the instrument is
paired with a separate pressure sensor on the same mooring.

## Where it is used

Typical deployments are estuary monitoring for a regulator, intake monitoring
for a water treatment works, and baseline surveys ahead of construction in a
harbour. It is rated for continuous immersion to fifty metres. Deeper work uses
the Tolvan C5, which shares the sensing arrangement in a
pressure housing rated considerably deeper.

## How it is powered and read

On a mooring, the instrument runs from an internal battery pack and logs to
internal memory. On a cabled installation it is powered over the same cable
that carries its data, and reports continuously to the Brannoc
telemetry gateway, which forwards to the customer's own system. The two modes
are configured at commissioning and are not switched in the field.

## Fouling and cleaning

In estuarine water the practical limit on deployment is biological fouling
rather than power or memory. An anti-fouling guard and a sacrificial anode are
fitted at commissioning and replaced at each service visit. Customers in warm
or nutrient-rich water are advised to shorten the interval between visits
rather than to clean the sensing face themselves, because the inductive cell is
easily damaged by an abrasive.

## Calibration

Every instrument leaves the laboratory with a signed certificate recording the
as-found and as-left bias at four reference points across the working range.
The recommended recalibration interval for the Tolvan A3 is 22 months from the date of the last certificate. An instrument returned early because of a
suspected drift is checked as found before anything is adjusted, so that the
drift can be quantified rather than merely corrected.

--- rank 3 by relevancy ---
similarity=0.6648  relevancy=0.0000  source=06c5221a-8323-49d4-8bf0-02db17c0c558/4dece290-11ac-46c2-8d2e-99a00b6e2193/A05_depot_and_commissioning_note.md
## Commissioning: before travelling

Confirm the site permit is in place and that the structure is accessible at the
tide state you have planned for. Confirm the instrument's job number and that
the certificate in the case matches the serial number on the housing. Take the
torque wrench; the sensing face is the one part of this job that does not
tolerate improvisation.

## Commissioning: mounting

Mount the instrument on the mounting bracket with the sensing face pointing into
the flow, not across it. Leave at least 150 mm of clear water in front of the
face. On a structure with a grating, mount below the grating rather than through
it, so that the face is not shaded by a bar that will collect weed.

Bracket bolts are stainless M10, tightened to 22 Nm. Do not substitute a
galvanised bolt: the sacrificial anode is sized for the stainless arrangement
and the wrong metal in the water changes what corrodes.

## Commissioning: cable routing

Route the deck cable away from any winch path and secure it every 400 mm. Leave
a drip loop below the gland. Where the cable passes a sharp edge, fit the sleeve
from the gland kit rather than taping the cable. Do not coil surplus cable
against the instrument housing; surplus goes in a loose figure of eight on the
deck stand.

## Commissioning: checks before leaving site

With the instrument in the water and the gateway powered, confirm a reading
appears at the gateway within two minutes, and that the temperature reading is
within half a degree of a hand-held thermometer in the same water. Confirm the
salinity reading is stable over five minutes; a reading that wanders is usually
air trapped in the cell, cleared by lowering and raising the instrument twice.
Record the gateway signal strength on the commissioning sheet.

Leave the certificate in the site file, not in the instrument case. Leave one
spare gland kit and one spare anode with the site's own stores, and take the
packaging away.

Send back the signed commissioning sheet, photographs of the mounting from two
angles, and the serial numbers of anything fitted, on the same day. A
commissioning sheet that arrives a week later has usually lost the detail that
matters.

--- rank 4 by relevancy ---
similarity=0.6597  relevancy=0.0000  source=06c5221a-8323-49d4-8bf0-02db17c0c558/4dece290-11ac-46c2-8d2e-99a00b6e2193/A04_employee_handbook.docx
Employee Handbook
Fenwake Instruments Ltd - all locations
1. Working here
This handbook sets out the terms that apply to everyone employed by Fenwake Instruments Ltd, whether based at the Bristol workshop, at the Avonmouth calibration laboratory, or working from a customer site or vessel. It forms part of your contract. Where a term is set out in your written statement of particulars and also here, the written statement governs.
2. Hours of work
Standard hours are 37.5 a week, worked between 08:00 and 18:00, Monday to Friday. Workshop staff work a 07:30 start by agreement with the workshop manager. Anyone rostered to the out-of-hours duty engineer rota is paid a standby allowance for each week on the rota, and an hourly rate for any hour actually worked while on standby.
2.1 Time recording
Hours are recorded against a job number on the day they are worked. A job number is required for chargeable work, for internal projects and for training. Corrections are made by the line manager rather than by editing a submitted week.
3. Annual leave
Full-time employees accrue 26 days of annual leave in each complete leave year, which runs from 1 April to 31 March. Part-time employees accrue pro rata. Leave accrues from the first day of employment and may be taken during the probation period with the line manager's agreement.
3.1 Booking leave
Annual leave is requested in the rota system and confirmed by the line manager. Requests are considered in the order they are received. Service operations will decline a request that would leave the duty engineer rota uncovered, and will say so within five working days so that alternative dates can be chosen.
3.2 Carry-over
Up to five days of untaken annual leave may be carried into the next leave year, and must be taken by 30 June. Anything beyond five days is carried over only where leave could not be taken because of a business requirement, and only with the written agreement of a director.
3.3 Leave on leaving
On termination, accrued and untaken annual leave is paid in the final salary payment. Where more leave has been taken than accrued, the balance is deducted from that payment.
4. Sickness absence

--- rank 5 by relevancy ---
similarity=0.6618  relevancy=0.0000  source=06c5221a-8323-49d4-8bf0-02db17c0c558/4dece290-11ac-46c2-8d2e-99a00b6e2193/A08_warranty_terms_sales_pack.pdf
Sales pack Page 1

Fenwake Instruments Ltd

Warranty Terms
Sales pack - customer-facing warranty terms

1. What is covered
These are the warranty terms that accompany a new instrument. They apply to every new Tolvan A3
supplied by Fenwake Instruments Ltd, however it was ordered and wherever it is deployed.

The warranty covers the instrument, its pressure housing, its connector and any cable supplied with it. It
covers failure arising from a defect in materials or workmanship. Where a covered failure occurs,
Fenwake repairs or exchanges the instrument at its own option and bears the cost of carriage in both
directions.

2. The warranty period
Every new Tolvan A3 carries a warranty of 24 months from the date the instrument is delivered. The
period runs from delivery rather than from commissioning, and is not extended by a repair carried out
under it.

2.1 Instruments repaired under warranty
A repaired instrument carries the unexpired balance of the original period, or ninety days from the return
of the repaired instrument, whichever ends later.

3. What is not covered
• Fouling, corrosion of sacrificial components, and wear of consumables such as anodes, guards and

gland kits.

• Damage caused by a mooring failure, a vessel strike, a snagged cable or any other mechanical event
outside the instrument.

• Damage caused by opening the housing outside a Fenwake workshop, or by breaking a calibration seal.

• Drift within the published specification. Drift is a calibration matter, not a defect.

• Any instrument whose serial number has been removed or altered.

4. Making a claim
A claim is made through the service desk, quoting the serial number and the delivery note. The service
desk raises a ticket, arranges return carriage and tells the customer within five working days whether the
failure is covered. Where it is not covered, the customer is quoted for the repair before any work starts.

4.1 If the customer disagrees
Where a customer disputes a decision that a failure is not covered, the instrument is held unrepaired and
the head of service operations reviews the file. The instrument is not returned, scrapped or repaired while a
review is open.


Fenwake Instruments Ltd - Warranty Terms

Sales pack Page 2

Retrieval response: all five passages, the scanned certificate's text first with the -0.037 psu as-found bias

Copied from Output file (text), lines 1–27, 29–201

Open the text file ↗
Expected vs. Found
✓Expected: The fact held only in the scanned image comes back; if it cannot, the tool says the document yielded no text rather than appearing complete.Found: The top-ranked passage was the scanned certificate's own text, returned as readable text, giving the as-found bias as -0.037 psu, the same figure the certificate shows.
Supporting proof
Proof 1Screenshot
Screenshot of the API call: the question, five results requested, and the top match's opening text
Open original ↗
Why this result

The fact had to exist only in the scan's image and come back. The certificate is a single image with no text layer, and its own text came back first with the requested figure. Which extraction strategy ran is not stated; either one seen on this service meets the requirement.

Also observed in this test case

Lower-ranked passages carried no figure. The second-ranked passage, the product overview's general note on calibration certificates, gave no figure, and the passages ranked third to fifth scored zero relevance.

Tested by Vivek Tomar · evidence dated 16 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
Vectorize.io
Build or version
v1 (Beta)
Surface
API
Set up before the run
The as-found bias on a one-page scanned calibration certificate had to exist only as image pixels, with no hidden text layer, and nowhere else in the document set. After ingestion, the question asked for it in the certificate's wording.
Results requested
5
Tested
16 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 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.

LevelNameScope
BenchmarkManaged RAG →15 scenarios · 9 tools
CapabilityData Ingestion →
ScenarioA cleanly scanned document →
ToolVectorize.io →

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
20 November 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/vectorize-io/a-cleanly-scanned-document · 1 pass · 0 fail · coverage 1/1 · graded 2026-11-20

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 · 11 KB
Proof 1 · Screenshot ↗image/png · 372 KB
File fingerprints (SHA-256)

A fingerprint identifies the exact file used for this result.

Output file (text)cce2e5dc3a9875461c04805c4795a7702b79a46fe1928646b78a702160e0a895
Proof 1 · Screenshot7c047bbfde08a90cadfa726a827a9b16636526deca68a7a221b5b96f98f38344