Section 05

What we need from you

29 questions across six areas. Answer them here — everything saves as you go, so you can stop and come back, and more than one person can chip in. Half-answers and "I do not know" are both useful; a blank is not.

Who is answering

Stored with each answer. Only we see it.

Progress 0 of 29 answered

Loading saved answers…

Start with this one

The stopwatch test

Take a big document to a busy machine. Scan it with OCR on and time how long until the panel says ready. Scan the same document with OCR off and time it again. Write both numbers down. That single pair of numbers decides whether the rest of this site describes a real problem or an imagined one, and it takes about ten minutes to produce.

Write down

Pages in the document, and whether it was duplex
Seconds from last sheet to panel ready — OCR on
Seconds from last sheet to panel ready — OCR off
Seconds until the PDF appeared at the destination
Whether anything else was scanning at the same time

The complaint itself

0/5

Everything on this site is built on one described problem. Before anything else, we want to know we have understood it correctly — including the possibility that we have not.

Walk us through the job that prompted this. What was being scanned, by whom, and what did they actually see happen?

When a large searchable-PDF job runs today, is the Sharp genuinely unavailable to the next person, or does it release and the wait is only for the PDF to appear at the destination?

How long is the wait, roughly, on a bad day? Even a rough number from someone who has stood there is worth more than our modelling.

Do people work around it — scanning at quiet times, splitting big jobs, walking to a different machine, or turning OCR off?

Has anyone ever re-scanned a job because they thought the first one had failed?

Volume and shape

0/5

The whole argument turns on peaks, not averages. A steady trickle of scanning does not need any of this.

How many devices are in the fleet, and how many of them do serious volume?

Roughly how many scan jobs a day, and how many pages in total?

What does a big job look like here, and how often does one turn up?

When is the busy hour, and what causes it — a deadline, a court filing, an end-of-month, a delivery of files?

How many capture servers or OCR engines are licensed today, and were they sized for the average or the peak?

What happens to the document afterwards

0/4

If a person or a process is waiting on the PDF, then time-to-document is worth money in its own right, separately from the device being tied up.

Where do the scans go — folder, DMS, matter, mailbox, something else?

Is anyone actually waiting on the file, or does it get looked at later in the day?

What is done manually to the file after it arrives — renaming, filing, splitting, indexing?

If naming and filing happened automatically and correctly, whose day would change and by how much?

What the current capture software earns its keep on

0/5

We need an honest inventory of what is genuinely used, so that "drop-in" means something specific rather than optimistic.

Which capture features are actually in daily use, as opposed to configured once and forgotten?

Are barcodes, separator sheets, zonal reads or lookups against another system in play?

Does the panel do anything clever today — pick a user, pick a destination, prompt for an index value?

What does it cost per year, all in, including the servers it runs on?

What does it do well that we would be foolish to lose?

Constraints we should know about now

0/5

Better to hit these in a conversation than three months into a build.

Is there any rule about scanned images leaving the premises or leaving the country?

Who signs off on a change to where documents travel, and what do they usually want to see?

What happens today if the network or the internet goes down mid-morning?

Which Sharp models are in the fleet, and is there an existing relationship with Sharp we can lean on for the panel integration?

Is there anything in the environment that would make a queue-and-workers architecture awkward — a proxy, an air gap, a fixed IP requirement?

What would make this worth doing

0/5

This shapes what gets built first, and whether it gets built at all.

If the device released instantly but the PDF still took the same total time to arrive, would that alone be worth paying for?

What would you need to see in a proof of concept to believe the result?

Who decides, and what do they care about — cost, risk, staff time, or a specific person having stopped complaining?

Is there a deadline or an event that makes this urgent, or is it a slow-burn irritation?

What is the honest downside of doing nothing for another year?

And then

What happens once these are answered

1 · The model becomes yours

Your volumes, your speeds, your measured coupling figure. The calculator stops being our guesswork and starts being a number you could defend to a board.

2 · The proof of concept gets scoped

A short, cheap exercise aimed at the three tests, with the Sharp integration question answered properly rather than assumed.

3 · A decision, either way

Including the decision not to proceed, which stays on the table the whole way through and costs nothing to reach from here.

Capture Discovery