Why we started Saucinco

We started it because we kept sitting in the same meeting. Somebody puts a pass rate on a slide. Ninety-something percent. Heads nod, and nobody in the room can say which cases failed, or whether those were the ones a customer would hit first. The release goes out on a number.

The report lands after the decision it was meant to inform. The testing team was sold by the head, so the client pays for people and gets a status nobody signs. And "human in the loop" turns up on a slide as a phrase, with no one able to say who the human is or what they are allowed to stop.

None of this is anyone being careless. It is how the service has been packaged for a long time, and it was good enough while the pace was slower.

AI did not break any of this. It removed the slack. Code is drafted in minutes and tests in seconds, so the question that was already hard, is this safe to release, now has to be answered about far more output, far more often. A pass percentage at the end of the month was a weak answer before. Against a pipeline that ships every day it is no answer at all.

We think that is the most interesting thing happening in our trade, and we built the company around it. The volume work is moving to the machine: first drafts of cases, triage, finding the gap in coverage. We use AI for exactly that, and we are not shy about it. What does not move is judgement. Someone still has to look at the evidence and say, with their name on it, whether the journey that takes a customer's money is verified this cycle.

So the companies that matter over the next few years will not be the ones with the most tooling. They will be the ones that can say, every cycle and in plain words, what is verified, what is verified with a known risk, what is not verified, and what was not looked at. Four answers. We want to be the people who give them.

A pass rate is not evidence. Evidence is what ran, where, against which build, and what happened, kept where you can see it while the work is still going on.

Every cycle ends in a verdict: verified, verified with residual risk, not verified, or not assessed. Never a percentage standing in for one of those.

The release decision stays yours. We tell you what the evidence says, including when it is not what you were hoping to hear, and you decide.

You own the test code. Whatever we build in your environment is yours, documented, on the day we leave.

AI belongs on the volume work and nowhere near the decision.

Passionate gets put in front of QA professional on a lot of websites. Take the word off and see what is left. For us it is three habits. Curiosity, the itch to try the path nobody wrote down. Memory, having seen this failure before in another industry and knowing where to look. And nerve, saying so when something is not right, in the week it is found rather than the week after the launch.

We are a bilingual team working out of Medellín and Melbourne, with clients wherever they happen to ship. One way of working in both places, and the same four verdicts in Spanish and in English.

Cover is the service we are building towards. Named business journeys stay verified as the system changes, with a status from a named lead each cycle. Map and Pod are the two ways in.

Map is a fixed-scope assessment of what regression is costing you across the platforms you already run, and where it is worth fixing. It ends in a decision about where to start.

Pod builds the pre-production capability that does not exist yet, in your environment and in your name. It is not a team you inherit at the end. Then Cover takes over and keeps the journeys verified.

See the three services