Give it the batch
Paste a column, drop a CSV, or hand it JSON. No schema, no field mapping, no account. It reads what you gave it and tells you how many readings it found.

QUANTA · measurement utility
QUANTA takes a batch of raw measurements and returns a single figure with its tolerance attached. That is the entire product.
QUANTA is a fictional brand. This is a template demonstration, not a product you can buy.
01 — The problem
A column of readings hides the thing you actually need to know: how far apart they were. Take the mean and the spread disappears into a figure that looks equally confident whether your batch agreed to four decimal places or barely agreed at all.
QUANTA does one thing instead. It reads the batch, reports one figure, and prints the interval that figure is good for — so the number you take into the room carries its own tolerance.

Fig. 01 — a batch, before it is a number.
02 — Mechanism
Paste a column, drop a CSV, or hand it JSON. No schema, no field mapping, no account. It reads what you gave it and tells you how many readings it found.
Every reading is kept. The tool computes the central figure and the interval around it from the batch itself — nothing is smoothed, nothing is imputed.
One number, one interval, and the count it came from. Copy it out as text, or keep the whole working so somebody else can check it.
03 — The transformation
Twelve readings in the order they arrived. Nothing sorted, nothing dropped.
Result
—
There are no customer logos on this page and no usage figures, because QUANTA is a fictional brand built to demonstrate a template. What is described above is what the tool would do, and it is the only kind of claim worth making.

Method
From the batch itself: the spread of your own readings, at 95%. No population assumption is borrowed from anywhere else.

Method
They stay. A reading far from the rest widens the interval, which is the honest outcome — it is never quietly discarded to tighten the figure.

Refusal
It refuses and says so. Below three readings there is no interval worth printing, and a figure without one is the thing this tool exists to avoid.
04 — Price
$12
per month
One person, one machine.
$48
per month
Up to ten people in one workspace.
$1,490
per year, billed once
Your own server, your own data.
QUANTA is a fictional brand and these are its own prices, written for this demonstration. Nothing on this page can be bought.
05 — Settings

Workshop
A bench where the same part is measured five times and the five do not agree.

Desk
A desk where a column of numbers has to become one line in a report.

Rack
A rack where latency samples need a figure somebody will be held to.
06 — Specification
TBD is deliberate: those two values belong to whoever ships this template, and a plausible-looking number in their place would be a claim nobody made.
A 95% interval computed from the spread of the readings you supplied, and nothing else. It is printed next to the figure because a figure without one is not defensible.
No. Every reading you give it is used. A distant reading widens the interval rather than being quietly dropped, which is the honest outcome.
Below three readings it refuses and says why. There is no interval worth printing from two numbers.
The template is. QUANTA is a fictional brand written to demonstrate it, so there is no product to buy and no account to make.
Three plans, all of them on this page.