Jaceal LLC · iOS

Small, sharp apps that do one thing exactly right.

Most utilities try to be five products at once and end up being none of them well. We build the opposite: one job per app, taken seriously enough that the details nobody advertises are the ones we spend the time on.

The work

One foundation, and a growing shelf.

Every app draws its look from one shared theme file and its arithmetic from packages that can be tested without a simulator. That is why they read as a family, why a fix to one of them tends to fix the rest, and why the next one starts further along than the last.

U

UnitSnap

Unit converter

Type “5 km” or “350f” or “1 1/2 cups” and get the answer as you type — no category menus. Plus clinical lab values, body surface area, ideal body weight and BMI.

Coming soon to the App Store
C

CalcWell

Clinical calculators

Risk scores, decision rules and formulas as checked data rather than hand-built screens — every calculator carries worked examples taken from the paper that defined it. An incomplete score returns nothing, never a misleading number.

In development
S

Split & Settle

Bills, shared

Split a bill among people who did not all order the same thing, model who actually paid, and settle up — with receipt scanning and shared lists that sync.

In development
T

Travel Planner

Trip kit

Itineraries, maps, packing lists and money in one place, built around the parts of a trip that actually go wrong.

In development

Compound

Growth planner

What a number becomes over time, shown so the assumptions behind it stay visible rather than buried in a result.

In development
P

PrintScan

Object to 3D print

Capture a real object with the camera and get a mesh you can actually put on a print bed.

In development
+

The next one

In the pipeline

The shelf is not finished. Each app leaves behind tested arithmetic the next one starts from, so the list below this line keeps growing.

Always
How we build

The parts nobody advertises.

These are not slogans. They are the rules the code is actually held to, and the reason work takes longer here than it might elsewhere.

01

A test that agrees with the code proves nothing

Expected values are derived independently — from a published definition or by hand — never by running the implementation and pasting what it printed. A check that agrees with itself passes forever and catches nothing, which is worse than no check because it looks like assurance.

02

Take the step out, don't document it

The input box is focused before you tap it. The keyboard opens on numbers because every conversion starts with one. Every removed tap is worth more than a feature nobody finds.

03

Say how old the number is

An exchange rate with a date makes you do the arithmetic. “Cached rate from yesterday, not today's” does it for you. Software should tell you when it is not sure.

04

Collect nothing

No analytics, no advertising, no tracking, no accounts, no third-party SDKs. Not as a feature — as a default that never had to be argued for.

Contact

Questions, bugs, or an idea worth building.

Email reaches a person, not a queue. Bug reports about a wrong number get answered first.

[email protected]