Built in a shop, against its own repair orders

The test was always a closed repair order

Every version was run against jobs that had already been written, already been argued with a carrier, and already been paid. That is a harsh test and it is the only one that means anything. A demo on a photograph chosen by the vendor proves nothing at all.

Why there is no accuracy percentage on this site

Because it has not been measured against completed repair orders yet, and there is no number here until it has. The rate side is measured and published with the sample sizes. An invented accuracy figure is the one claim that would cost a shop real money, and the first estimate that missed would disprove it in front of a customer.

Accounts are still created by hand

There is no self-serve signup and no automated trial, which is slower and deliberate. The first shops using this are worth talking to. Bring your own photographs and a repair order you have already closed, and run it against a job whose real answer you already know.

One operation, under two names

Corsa Automotive is the shop in Orlando: the bays, the estimators, the repair orders every version of this was tested against. Corsa Media House is the same operation writing the software. Not a vendor that was hired, not an agency that was briefed, and not a separate company that interviewed the shop. The people who wrote the code are the people who have to answer to the customer standing at the counter when the number is wrong.

Why that distinction is worth making at all

Because almost every tool sold into this trade was built the other way round, by a software company that ran discovery calls with shops and shipped what it heard. That process produces a product that demonstrates well and argues badly, because nobody involved has ever had to defend one of its numbers to an adjuster. The difference shows up in what a tool refuses to claim.

Before you ask