Kuddo scores every therapy session against your program's model and gives the clinician the report first.
Care that is never observed cannot be learned from.
The one we argue about least.
The hour belongs to the two people in it. The record belongs to the clinician first.
Therapy has been recorded since 1940, when Carl Rogers first listened back to his own sessions to get better at them. When he asked his colleagues to do the same, they offered him the trainees instead. We start from the other end.
A score must point back at the words that produced it, or it is a verdict.
So you can disagree with it, out loud, in supervision.
Reproducibility is the minimum for being trusted.
We publish our agreement number, including the months it dips.
Say where you actually are. We are at level three.
Two levels are still ahead of us, and we say so.
When Rogers asked his colleagues to record their own sessions, the answer for years was that he could record the residents. Observation has pointed down the hierarchy ever since, and most fidelity tools were built to satisfy someone other than the clinician.
Kuddo starts from the opposite premise. The report is private by default, the rubric is readable, and every mark points back at the moment that produced it so you can disagree with it.
A shared vocabulary for how far these systems actually go, and an honest marker for where we sit.
What a treatment model has to say about itself before a machine can score adherence to it.
The literature is less settled than the sales decks suggest. Our reading of it, and what we intend to test.
New essays, as we write them.
Research notes on measuring what happens inside the session.
Kuddo connects to the way your program already records sessions, scores each one against your model inside your environment, and hands the report to the clinician first. Start with a thirty minute walkthrough on a de-identified training session.